Vous êtes sur la page 1sur 766

Nuclear Medicine Physics

A Handbook for Teachers and Students




D.L. Bailey
J.L. Humm
A. Todd-Pokropek
A. van Aswegen
Technical Editors

Nuclear Medicine Physics:

A Handbook for Teachers and

The following states are Members of the international atomic energy agency:
bosNia aNd herZeGoViNa
bruNei darussalaM
burkiNa faso
ceNTral africaN
cosTa rica
cTe diVoire
cZech rePublic
deMocraTic rePublic
of The coNGo
doMiNicaN rePublic
el salVador

holy see
iraN, islaMic rePublic of
korea, rePublic of
lao PeoPles deMocraTic
Marshall islaNds
MauriTaNia, islaMic
rePublic of
NeW ZealaNd

PaPua NeW GuiNea
rePublic of MoldoVa
russiaN federaTioN
saN MariNo
saudi arabia
sierra leoNe
souTh africa
sri laNka
syriaN arab rePublic
The forMer yuGoslaV
rePublic of MacedoNia
TriNidad aNd TobaGo
uNiTed arab eMiraTes
uNiTed kiNGdoM of
GreaT briTaiN aNd
NorTherN irelaNd
uNiTed rePublic
of TaNZaNia
uNiTed sTaTes of aMerica
VeNeZuela, boliVariaN
rePublic of
VieT NaM

The agencys statute was approved on 23 october 1956 by the conference on the statute of the
iaea held at united Nations headquarters, New york; it entered into force on 29 July 1957. The
headquarters of the agency are situated in Vienna. its principal objective is to accelerate and enlarge the
contribution of atomic energy to peace, health and prosperity throughout the world.

Nuclear Medicine Physics:

A Handbook for Teachers and
endorsed by:

International atomic energy agency

Vienna, 2014

All IAEA scientific and technical publications are protected by the terms of
the Universal Copyright Convention as adopted in 1952 (Berne) and as revised
in 1972 (Paris). The copyright has since been extended by the World Intellectual
Property Organization (Geneva) to include electronic and virtual intellectual
property. Permission to use whole or parts of texts contained in IAEA publications
in printed or electronic form must be obtained and is usually subject to royalty
agreements. Proposals for non-commercial reproductions and translations are
welcomed and considered on a case-by-case basis. Enquiries should be addressed
to the IAEA Publishing Section at:
Marketing and Sales Unit, Publishing Section
International Atomic Energy Agency
Vienna International Centre
PO Box 100
1400 Vienna, Austria
fax: +43 1 2600 29302
tel.: +43 1 2600 22417
email: sales.publications@iaea.org
IAEA, 2014
Printed by the IAEA in Austria
December 2014

IAEA Library Cataloguing in Publication Data

Nuclear medicine physics : a handbook for students and teachers. Vienna :
International Atomic Energy Agency, 2014.
p. ; 24 cm.
ISBN 9789201438102
Includes bibliographical references. 1. Nuclear medicine Handbooks, manuals, etc.
2. Medical physics handbooks. 3. Medical physics. I. International Atomic Energy

Nuclear medicine is the use of radionuclides in medicine for diagnosis,
staging of disease, therapy and monitoring the response of a disease process.
It is also a powerful translational tool in the basic sciences, such as biology, in
drug discovery and in pre-clinical medicine. Developments in nuclear medicine
are driven by advances in this multidisciplinary science that includes physics,
chemistry, computing, mathematics, pharmacology and biology.
This handbook comprehensively covers the physics of nuclear medicine.
It is intended for undergraduate and postgraduate students of medical physics.
It will also serve as a resource for interested readers from other disciplines, for
example, clinicians, radiochemists and medical technologists who would like to
familiarize themselves with the basic concepts and practice of nuclear medicine
The scope of the book is intentionally broad. Physics is a vital aspect of
nearly every area of nuclear medicine, including imaging instrumentation,
image processing and reconstruction, data analysis, radionuclide production,
radionuclide therapy, radiopharmacy, radiation protection and biology. The
authors were drawn from a variety of regions and were selected because of their
knowledge, teaching experience and scientific acumen.
This book was written to address an urgent need for a comprehensive,
contemporary text on the physics of nuclear medicine. It complements similar
texts in radiation oncology physics and diagnostic radiology physics that have
been published by the IAEA.
Endorsement of this handbook has been granted by the following
international professional bodies: the American Association of Physicists in
Medicine (AAPM), the AsiaOceania Federation of Organizations for Medical
Physics (AFOMP), the Australasian College of Physical Scientists and Engineers
in Medicine (ACPSEM), the European Federation of Organisations for Medical
Physics (EFOMP), the Federation of African Medical Physics Organisations
(FAMPO), and the World Federation of Nuclear Medicine and Biology
The following international experts are gratefully acknowledged
for making major contributions to this handbook as technical editors:
D.L. Bailey (Australia), J.L. Humm (United States of America), A. Todd-Pokropek
(United Kingdom) and A. van Aswegen (South Africa). The IAEA officers
responsible for this publication were S. Palm and G.L. Poli of the Division of
Human Health.

Although great care has been taken to maintain the accuracy of information contained
in this publication, neither the IAEA nor its Member States assume any responsibility for
consequences which may arise from its use.
The use of particular designations of countries or territories does not imply any
judgement by the publisher, the IAEA, as to the legal status of such countries or territories, of
their authorities and institutions or of the delimitation of their boundaries.
The mention of names of specific companies or products (whether or not indicated as
registered) does not imply any intention to infringe proprietary rights, nor should it be construed
as an endorsement or recommendation on the part of the IAEA.
The IAEA has no responsibility for the persistence or accuracy of URLs for external or
third party Internet web sites referred to in this book and does not guarantee that any content
on such web sites is, or will remain, accurate or appropriate.

Nuclear medicine is the study and utilization of radioactive compounds in
medicine to image and treat human disease. It relies on the tracer principle first
espoused by Georg Karl von Hevesy in the early 1920s. The tracer principle is
the study of the fate of compounds in vivo usingminute amounts of radioactive
tracers which do not elicit any pharmacological response by the body to the tracer.
Today, the same principle is used to study many aspects of physiology, such as
cellular metabolism, DNA (deoxyribonucleic acid) proliferation, blood flow in
organs, organ function, receptor expression and abnormal physiology, externally
using sensitive imaging devices. Larger amounts of radionuclides are also applied
to treat patients with radionuclide therapy, especially in disseminated diseases
such as advanced metastatic cancer, as this form of therapy has the ability to
target abnormal cells to treat the disease anywhere in the body.
Nuclear medicine relies on function. For this reason, it is referred to as
functional imaging. Rather than just imaging a portion of the body believed
to have some abnormality, as is done with X ray imaging in radiology, nuclear
medicine scans often depict the whole body distribution of the radioactive
compound often acquired as a sequence of images over time showing the
temporal course of the radiotracer in the body.
There are two main types of radiation of interest for imaging in nuclear
medicine: ray emission from excited nuclei, and annihilation (or coincidence)
radiation () arising after positron emission from proton-rich nuclei. Gamma
photons are detected with a gamma camera as either planar (2D) images or
tomographically in 3D using single photon emission computed tomography.
The annihilation photons from positron emission are detected using a positron
emission tomography (PET) camera. The most recent major development in this
field is the combination of gamma cameras or PET cameras with high resolution
structural imaging devices, either X ray computed tomography (CT) scanners
or, increasingly, magnetic resonance imaging (MRI) scanners, in a single image
device. The combined PET/CT (or PET/MRI) scanner represents one of the most
sophisticated and powerful ways to visualize normal and altered physiology in
the body.
It is in this complex environment that the medical physicist, along with
nuclear medicine physicians and technologists/radiographers, plays a significant
role in the multidisciplinary team needed for medical diagnosis. The physicist is
responsible for such areas as instrumentation performance, radiation dosimetry
for treatment of patients, radiation protection of staff and accuracy of the data
analysis. The physicist draws on training in radiation and nuclear science,
in addition to scientific rigour and attention to detail in experiments and
measurements, to join forces with the other members of the multidisciplinary

team in delivering optimal health care. Patients are frequently treated on the
basis of the result of the scans they receive and these, therefore, have to be of the
highest quality.
This handbook was conceived and written by physicists, and is intended
primarily for physicists, although interested readers from medical, paramedical
and other science and engineering backgrounds could find it useful. The level
of understanding of the material covered will be different depending on the
background of the reader. Readers are encouraged to visit the IAEA Human
Health web site (http://www-naweb.iaea.org/NAHU/index.html) to discover the
wealth of resources available.
The technical editors and authors, selected for their experience and in
recognition of their contributions to the field, were drawn from around the world
and, thus, this book represents a truly international collaboration. The technical
editors travelled to the IAEA headquarters in Vienna on four occasions over three
years to bring this project to fruition. We would like to thank all of the authors for
their important contribution.
D.L. Bailey, J.L. Humm
A. Todd-Pokropek, A. van Aswegen





INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1.1. Fundamental physical constants . . . . . . . . . . . . . . . . . . . 1
1.1.2. Physical quantities and units. . . . . . . . . . . . . . . . . . . . . . 2
1.1.3. Classification of radiation. . . . . . . . . . . . . . . . . . . . . . . . 4
1.1.4. Classification of ionizing radiation. . . . . . . . . . . . . . . . . 4
1.1.5. Classification of indirectly ionizing photon radiation. . . 5
1.1.6. Characteristic X rays. . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.1.7. Bremsstrahlung . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.1.8. Gamma rays. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
1.1.9. Annihilation quanta . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
1.1.10. Radiation quantities and units. . . . . . . . . . . . . . . . . . . . . 7
1.2.1. Rutherford model of the atom. . . . . . . . . . . . . . . . . . . . . 10
1.2.2. Bohr model of the hydrogen atom. . . . . . . . . . . . . . . . . . 10
1.3.1. Nuclear radius. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
1.3.2. Nuclear binding energy. . . . . . . . . . . . . . . . . . . . . . . . . . 12
1.3.3. Nuclear fusion and fission. . . . . . . . . . . . . . . . . . . . . . . . 13
1.3.4. Two-particle collisions and nuclear reactions. . . . . . . . . 14
RADIOACTIVITY. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
1.4.1. Decay of radioactive parent into a stable or unstable
daughter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
1.4.2. Radioactive series decay . . . . . . . . . . . . . . . . . . . . . . . . 19
1.4.3. Equilibrium in parentdaughter activities. . . . . . . . . . . . 21
1.4.4. Production of radionuclides (nuclear activation) . . . . . . 22
1.4.5. Modes of radioactive decay . . . . . . . . . . . . . . . . . . . . . . 23
1.4.6. Alpha decay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
1.4.7. Beta minus decay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
1.4.8. Beta plus decay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
1.4.9. Electron capture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
1.4.10. Gamma decay and internal conversion . . . . . . . . . . . . . 27
1.4.11. Characteristic (fluorescence) X rays and
Augerelectrons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
1.5.1. Electronorbital interactions. . . . . . . . . . . . . . . . . . . . . . 29
1.5.2. Electronnucleus interactions. . . . . . . . . . . . . . . . . . . . . 29


PHOTON INTERACTIONS WITH MATTER. . . . . . . . . . . . . . 30

1.6.1. Exponential absorption of photon beam in absorber. . . . 30
1.6.2. Characteristic absorber thicknesses . . . . . . . . . . . . . . . . 31
1.6.3. Attenuation coefficients . . . . . . . . . . . . . . . . . . . . . . . . . 34
1.6.4. Photon interactions on the microscopic scale. . . . . . . . . 35
1.6.5. Photoelectric effect. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
1.6.6. Rayleigh (coherent) scattering. . . . . . . . . . . . . . . . . . . . . 39
1.6.7. Compton effect (incoherent scattering). . . . . . . . . . . . . . 39
1.6.8. Pair production . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
1.6.9. Relative predominance of individual effects . . . . . . . . . 46
1.6.10. Macroscopic attenuation coefficients . . . . . . . . . . . . . . . 47
1.6.11. Effects following photon interactions with absorber
and summary of photon interactions . . . . . . . . . . . . . . . 48

CHAPTER 2. BASIC RADIOBIOLOGY. . . . . . . . . . . . . . . . . . . . . . . . . . . 49



INTRODUCTION. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
RADIATION EFFECTS AND TIMESCALES. . . . . . . . . . . . . . 49
2.3.1. Types of ionizing radiation. . . . . . . . . . . . . . . . . . . . . . . 51
MODIFIERS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
2.4.1. Role of oxygen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
2.4.2. Bystander effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
DNA DAMAGE AND REPAIR. . . . . . . . . . . . . . . . . . . . . . . . . . 55
2.5.1. DNA damage. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
2.5.2. DNA repair . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
CELLULAR EFFECTS OF RADIATION . . . . . . . . . . . . . . . . . 56
2.6.1. Concept of cell death. . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
2.6.2. Cell survival curves . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
2.6.3. Dose deposition characteristics: linear energy transfer. . 57
2.6.4. Determination of relative biological effectiveness. . . . . 58
2.6.5. The dose rate effect and the concept of repeat
treatments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
2.6.6. The basic linearquadratic model. . . . . . . . . . . . . . . . . . 63
2.6.7. Modification to the linearquadratic model for
radionuclide therapies. . . . . . . . . . . . . . . . . . . . . . . . . . . 64
2.6.8. Quantitative intercomparison of different treatment
types. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
2.6.9. Cellular recovery processes . . . . . . . . . . . . . . . . . . . . . . 65
2.6.10. Consequence of radionuclide heterogeneity. . . . . . . . . . 66




TISSUES/ORGANS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
2.7.1. Classification of radiation damage (early versus late) . . 66
2.7.2. Determinants of tumour response. . . . . . . . . . . . . . . . . . 67
2.7.3. The concept of therapeutic index in radiation
therapy and radionuclide therapy . . . . . . . . . . . . . . . . . . 68
2.7.4. Long term concerns: stochastic and deterministic
effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
TARGETED RADIONUCLIDE THERAPY. . . . . . . . . . . . . . . . 69
2.8.1. Radionuclide targeting . . . . . . . . . . . . . . . . . . . . . . . . . . 69
2.8.2. Whole body irradiation. . . . . . . . . . . . . . . . . . . . . . . . . . 69
2.8.3. Critical normal tissues for radiation and
radionuclide therapies. . . . . . . . . . . . . . . . . . . . . . . . . . . 70
2.8.4. Imaging the radiobiology of tumours . . . . . . . . . . . . . . . 71
2.8.5. Choice of radionuclide to maximize therapeutic index. . 71

CHAPTER 3. RADIATION PROTECTION. . . . . . . . . . . . . . . . . . . . . . . . . 73




INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
3.2.1. The International Commission on Radiological
Protection system of radiological protection. . . . . . . . . . 74
3.2.2. Safety standards. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
3.2.3. Radiation protection quantities and units . . . . . . . . . . . . 77
A NUCLEAR MEDICINE FACILITY . . . . . . . . . . . . . . . . . . . . 81
3.3.1. General aspects. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
3.3.2. Responsibilities. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
3.3.3. Radiation protection programme. . . . . . . . . . . . . . . . . . . 84
3.3.4. Radiation protection committee . . . . . . . . . . . . . . . . . . . 84
3.3.5. Education and training. . . . . . . . . . . . . . . . . . . . . . . . . . . 84
FACILITY DESIGN. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
3.4.1. Location and general layout . . . . . . . . . . . . . . . . . . . . . . 85
3.4.2. General building requirements . . . . . . . . . . . . . . . . . . . . 85
3.4.3. Source security and storage . . . . . . . . . . . . . . . . . . . . . . 86
3.4.4. Structural shielding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
3.4.5. Classification of workplaces . . . . . . . . . . . . . . . . . . . . . 87
3.4.6. Workplace monitoring . . . . . . . . . . . . . . . . . . . . . . . . . . 88
3.4.7. Radioactive waste . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88






OCCUPATIONAL EXPOSURE . . . . . . . . . . . . . . . . . . . . . . . . . 89
3.5.1. Sources of exposure. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
3.5.2. Justification, optimization and dose limitation. . . . . . . . 91
3.5.3. Conditions for pregnant workers and young persons . . . 91
3.5.4. Protective clothing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
3.5.5. Safe working procedures. . . . . . . . . . . . . . . . . . . . . . . . . 92
3.5.6. Personal monitoring. . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
3.5.7. Monitoring of the workplace. . . . . . . . . . . . . . . . . . . . . . 95
3.5.8. Health surveillance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
3.5.9. Local rules and supervision. . . . . . . . . . . . . . . . . . . . . . . 96
PUBLIC EXPOSURE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
3.6.1. Justification, optimization and dose limitation . . . . . . . . 97
3.6.2. Design considerations. . . . . . . . . . . . . . . . . . . . . . . . . . . 97
3.6.3. Exposure from patients. . . . . . . . . . . . . . . . . . . . . . . . . . 98
3.6.4. Transport of sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
MEDICAL EXPOSURE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
3.7.1. Justification of medical exposure . . . . . . . . . . . . . . . . . . 99
3.7.2. Optimization of protection . . . . . . . . . . . . . . . . . . . . . . . 100
3.7.3. Helping in the care, support or comfort of patients. . . . . 107
3.7.4. Biomedical research . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
3.7.5. Local rules. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
POTENTIAL EXPOSURE. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
3.8.1. Safety assessment and accident prevention. . . . . . . . . . . 108
3.8.2. Emergency plans. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
3.8.3. Reporting and lessons learned. . . . . . . . . . . . . . . . . . . . . 111
QUALITY ASSURANCE. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
3.9.1. General considerations . . . . . . . . . . . . . . . . . . . . . . . . . . 112
3.9.2. Audit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114

CHAPTER 4. RADIONUCLIDE PRODUCTION. . . . . . . . . . . . . . . . . . . . 117



THE ORIGINS OF DIFFERENT NUCLEI. . . . . . . . . . . . . . . . . 117

4.1.1. Induced radioactivity. . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
4.1.2. Nuclide chart and line of nuclear stability. . . . . . . . . . . . 120
4.1.3. Binding energy, Q-value, reaction threshold and
nuclear reaction formalism . . . . . . . . . . . . . . . . . . . . . . . 123
4.1.4. Types of nuclear reaction, reaction channels and
cross-section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
REACTOR PRODUCTION. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
4.2.1. Principle of operation and neutron spectrum. . . . . . . . . 128



4.2.2. Thermal and fast neutron reactions . . . . . . . . . . . . . . . . 128

4.2.3. Nuclear fission, fission products . . . . . . . . . . . . . . . . . . 131
ACCELERATOR PRODUCTION. . . . . . . . . . . . . . . . . . . . . . . . 132
4.3.1. Cyclotron, principle of operation,
negativeandpositiveions. . . . . . . . . . . . . . . . . . . . . . . . 134
4.3.2. Commercial production (low and high energy). . . . . . . . 136
4.3.3. In-house low energy production (PET). . . . . . . . . . . . . . 137
4.3.4. Targetry, optimizing the production regarding yield
and impurities, yield calculations . . . . . . . . . . . . . . . . . . 140
RADIONUCLIDE GENERATORS. . . . . . . . . . . . . . . . . . . . . . . 141
4.4.1. Principles of generators. . . . . . . . . . . . . . . . . . . . . . . . . . 142
4.5.1. Carrier-free, carrier-added systems. . . . . . . . . . . . . . . . . 144
4.5.2. Separation methods, solvent extraction, ion
exchange, thermal diffusion . . . . . . . . . . . . . . . . . . . . . . 145
4.5.3. Radiation protection considerations and hot-box
facilities. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147





MEASUREMENT. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149
CHARACTERIZATION OF DATA. . . . . . . . . . . . . . . . . . . . . . . 153
5.2.1. Measures of central tendency and variability. . . . . . . . . 153
STATISTICAL MODELS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157
5.3.1. Conditions when binomial, Poisson and normal
distributions are applicable. . . . . . . . . . . . . . . . . . . . . . . 158
5.3.2. Binomial distribution. . . . . . . . . . . . . . . . . . . . . . . . . . . . 160
5.3.3. Poisson distribution. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
5.3.4. Normal distribution. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165
IMAGING . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168
5.4.1. Assumption. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168
5.4.2. The importance of the fractional F as an indicator
of the precision of a single measurement in sample
counting and imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . 170
5.4.3. Caution on the use of the estimate of the precision
of a single measurement in sample counting and
imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171




PROPAGATION OF ERROR . . . . . . . . . . . . . . . . . . . . . . . . . . . 172

5.5.1. Sums and differences. . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
5.5.2. Multiplication and division by a constant. . . . . . . . . . . . 174
5.5.3. Products and ratios . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176
5.6.1. Multiple independent counts. . . . . . . . . . . . . . . . . . . . . . 177
5.6.2. Standard deviation and relative standard deviation
for counting rates. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178
5.6.3. Effects of background counts . . . . . . . . . . . . . . . . . . . . . 179
5.6.4. Significance of differences between counting
measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
5.6.5. Minimum detectable counts, count rate and activity. . . . 184
5.6.6. Comparing counting systems . . . . . . . . . . . . . . . . . . . . . 187
5.6.7. Estimating required counting times. . . . . . . . . . . . . . . . . 188
5.6.8. Calculating uncertainties in the measurement of
plasma volume in patients. . . . . . . . . . . . . . . . . . . . . . . . 189
DETECTOR PERFORMANCE. . . . . . . . . . . . . . . . . . . . . . . . . . 191
5.7.1. Energy resolution of scintillation detectors. . . . . . . . . . . 191
5.7.2. Intervals between successive events. . . . . . . . . . . . . . . . 193
5.7.3. Paralysable dead time . . . . . . . . . . . . . . . . . . . . . . . . . . . 194

CHAPTER 6. BASIC RADIATION DETECTORS . . . . . . . . . . . . . . . . . . . 196



INTRODUCTION. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196
6.1.1. Radiation detectors complexity and relevance. . . . . . 196
6.1.2. Interaction mechanisms, signal formation and
detector type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196
6.1.3. Counting, current, integrating mode. . . . . . . . . . . . . . . . 197
6.1.4. Detector requirements. . . . . . . . . . . . . . . . . . . . . . . . . . . 197
GAS FILLED DETECTORS. . . . . . . . . . . . . . . . . . . . . . . . . . . . 200
6.2.1. Basic principles. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 200
SEMICONDUCTOR DETECTORS. . . . . . . . . . . . . . . . . . . . . . 202
6.3.1. Basic principles. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 202
6.3.2. Semiconductor detectors. . . . . . . . . . . . . . . . . . . . . . . . . 204
PHOSPHORS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205
6.4.1. Basic principles. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205
6.4.2. Light sensors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206
6.4.3. Scintillator materials. . . . . . . . . . . . . . . . . . . . . . . . . . . . 209


MEDICINE IMAGING DEVICES . . . . . . . . . . . . . . . . . . . . 214





INTRODUCTION. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214
7.2.1. Scintillation counters. . . . . . . . . . . . . . . . . . . . . . . . . . . . 215
7.2.2. Gas filled detection systems . . . . . . . . . . . . . . . . . . . . . . 216
7.2.3. Semiconductor detectors. . . . . . . . . . . . . . . . . . . . . . . . . 216
IMAGING DETECTORS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 217
7.3.1. The gamma camera. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 217
7.3.2. The positron camera . . . . . . . . . . . . . . . . . . . . . . . . . . . . 218
7.3.3. Multiwire proportional chamber based X ray and
ray imagers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
7.3.4. Semiconductor imagers. . . . . . . . . . . . . . . . . . . . . . . . . . 220
7.3.5. The autoradiography imager. . . . . . . . . . . . . . . . . . . . . . 221
SIGNAL AMPLIFICATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . 222
7.4.1. Typical amplifier . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 222
7.4.2. Properties of amplifiers. . . . . . . . . . . . . . . . . . . . . . . . . . 224
SIGNAL PROCESSING . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226
7.5.1. Analogue signal utilization. . . . . . . . . . . . . . . . . . . . . . . 226
7.5.2. Signal digitization. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226
7.5.3. Production and use of timing information. . . . . . . . . . . . 228
SYSTEMS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 230
7.6.1. Power supplies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 230
7.6.2. Uninterruptible power supplies. . . . . . . . . . . . . . . . . . . . 231
7.6.3. Oscilloscopes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231
SUMMARY. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232

CHAPTER 8. GENERIC PERFORMANCE MEASURES . . . . . . . . . . . . . 234



INTRINSIC AND EXTRINSIC MEASURES. . . . . . . . . . . . . . . 234

8.1.1. Generic nuclear medicine imagers . . . . . . . . . . . . . . . . . 234
8.1.2. Intrinsic performance. . . . . . . . . . . . . . . . . . . . . . . . . . . 236
8.1.3. Extrinsic performance. . . . . . . . . . . . . . . . . . . . . . . . . . . 236
ENERGY RESOLUTION. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237
8.2.1. Energy spectrum. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237
8.2.2. Intrinsic measurement energy resolution. . . . . . . . . . 238
8.2.3. Impact of energy resolution on extrinsic imager
performance. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239




SPATIAL RESOLUTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 240

8.3.1. Spatial resolution blurring. . . . . . . . . . . . . . . . . . . . . . . . 240
8.3.2. General measures of spatial resolution. . . . . . . . . . . . . . 241
8.3.3. Intrinsic measurement spatial resolution . . . . . . . . . . 242
8.3.4. Extrinsic measurement spatial resolution. . . . . . . . . . 242
TEMPORAL RESOLUTION. . . . . . . . . . . . . . . . . . . . . . . . . . . 244
8.4.1. Intrinsic measurement temporal resolution . . . . . . . . 244
8.4.2. Dead time. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 244
8.4.3. Count rate performance measures. . . . . . . . . . . . . . . . . . 246
SENSITIVITY. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247
8.5.1. Image noise and sensitivity. . . . . . . . . . . . . . . . . . . . . . . 247
8.5.2. Extrinsic measure sensitivity. . . . . . . . . . . . . . . . . . . 248
IMAGE QUALITY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 249
8.6.1. Image uniformity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 249
8.6.2. Resolution/noise trade-off. . . . . . . . . . . . . . . . . . . . . . . . 249
OTHER PERFORMANCE MEASURES. . . . . . . . . . . . . . . . . . 250

CHAPTER 9. PHYSICS IN THE RADIOPHARMACY. . . . . . . . . . . . . . . 251






9.1.1. Construction of dose calibrators. . . . . . . . . . . . . . . . . . . 251
9.1.2. Calibration of dose calibrators. . . . . . . . . . . . . . . . . . . . 253
9.1.3. Uncertainty of activity measurements. . . . . . . . . . . . . . . 254
9.1.4. Measuring pure emitters. . . . . . . . . . . . . . . . . . . . . . . . 258
9.1.5. Problems arising from radionuclide contaminants . . . . . 259
QUALITY CONTROL. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 260
9.2.1. Acceptance tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 260
9.2.2. Quality control. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 262
INDIVIDUAL PATIENTS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 264
9.5.1. Adjusting the activity for differences in patient size
and weight. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 264
9.5.2. Paediatric dosage charts . . . . . . . . . . . . . . . . . . . . . . . . . 264
9.5.3. Diagnostic reference levels in nuclear medicine. . . . . . . 266
9.6.1. Surface contamination limits. . . . . . . . . . . . . . . . . . . . . . 269
9.6.2. Wipe tests and daily surveys. . . . . . . . . . . . . . . . . . . . . . 270
9.6.3. Monitoring of staff finger doses during dispensing . . . . 270


PRODUCT CONTAINMENT ENCLOSURES. . . . . . . . . . . . . . 271

9.7.1. Fume cupboards. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 271
9.7.2. Laminar flow cabinets. . . . . . . . . . . . . . . . . . . . . . . . . . . 272
9.7.3. Isolator cabinets. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 273
9.8. SHIELDING FOR RADIONUCLIDES . . . . . . . . . . . . . . . . . . . 274
9.8.1. Shielding for , and positron emitters . . . . . . . . . . . . . 274
9.8.2. Transmission factors for lead and concrete. . . . . . . . . . . 278
9.9. DESIGNING A RADIOPHARMACY. . . . . . . . . . . . . . . . . . . . . 280
9.10. SECURITY OF THE RADIOPHARMACY. . . . . . . . . . . . . . . . 282
9.11. RECORD KEEPING. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 283
9.11.1. Quality control records . . . . . . . . . . . . . . . . . . . . . . . . . . 283
9.11.2. Records of receipt of radioactive materials. . . . . . . . . . . 283
9.11.3. Records of radiopharmaceutical preparation and
dispensing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 284
9.11.4. Radioactive waste records. . . . . . . . . . . . . . . . . . . . . . . . 284


10.1. INTRODUCTION. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 287
10.2.1. Ionization detectors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 288
10.2.2. Scintillation detectors . . . . . . . . . . . . . . . . . . . . . . . . . . . 292
10.3. RADIATION DETECTOR PERFORMANCE. . . . . . . . . . . . . . 294
10.3.1. Sensitivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 294
10.3.2. Energy resolution. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 295
10.3.3. Count rate performance (speed). . . . . . . . . . . . . . . . . . 296
10.4. DETECTION AND COUNTING DEVICES . . . . . . . . . . . . . . . 298
10.4.1. Survey meters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 298
10.4.2. Dose calibrator. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 299
10.4.3. Well counter. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 299
10.4.4. Intra-operative probes. . . . . . . . . . . . . . . . . . . . . . . . . . . 300
10.4.5. Organ uptake probe. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 302
COUNTING DEVICES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 305
10.5.1. Reference sources. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 305
10.5.2. Survey meter. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 306
10.5.3. Dose calibrator. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 307
10.5.4. Well counter. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 310
10.5.5. Intra-operative probe. . . . . . . . . . . . . . . . . . . . . . . . . . . . 310
10.5.6. Organ uptake probe. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 311


11.1. INTRODUCTION. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 312
11.2. GAMMA CAMERA SYSTEMS. . . . . . . . . . . . . . . . . . . . . . . . . 312
11.2.1. Basic principles. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 312
11.2.2. The Anger camera. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 314
11.2.3. SPECT. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 341
11.3. PET SYSTEMS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 353
11.3.1. Principle of annihilation coincidence detection . . . . . . . 353
11.3.2. Design considerations for PET systems . . . . . . . . . . . . . 356
11.3.3. Detector systems. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 362
11.3.4. Data acquisition. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 369
11.3.5. Data corrections. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 380
11.4. SPECT/CT AND PET/CT SYSTEMS. . . . . . . . . . . . . . . . . . . . . 392
11.4.1. CT uses in emission tomography . . . . . . . . . . . . . . . . . . 392
11.4.2. SPECT/CT. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 393
11.4.3. PET/CT. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 394
CHAPTER 12. COMPUTERS IN NUCLEAR MEDICINE. . . . . . . . . . . . . . 398
CAPABILITIES. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 398
12.1.1. Moores law. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 398
12.1.2. Hardware versus peopleware . . . . . . . . . . . . . . . . . . . . 398
12.1.3. Future trends. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 399
12.2. STORING IMAGES ON A COMPUTER. . . . . . . . . . . . . . . . . . 400
12.2.1. Number systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 400
12.2.2. Data representation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 401
12.2.3. Images and volumes . . . . . . . . . . . . . . . . . . . . . . . . . . . . 403
12.3. IMAGE PROCESSING. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 405
12.3.1. Spatial frequencies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 406
12.3.2. Sampling requirements. . . . . . . . . . . . . . . . . . . . . . . . . . 412
12.3.3. Convolution. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 412
12.3.4. Filtering. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 414
12.3.5. Band-pass filters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 416
12.3.6. Deconvolution. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 421
12.3.7. Image restoration filters. . . . . . . . . . . . . . . . . . . . . . . . . . 422
12.3.8. Other processing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 424
12.4. DATA ACQUISITION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 425
12.4.1. Acquisition matrix size and spatial resolution. . . . . . . . 426
12.4.2. Static and dynamic planar acquisition. . . . . . . . . . . . . . . 426

12.4.3. SPECT. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 427

12.4.4. PET acquisition. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 428
12.4.5. Gated acquisition. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 430
12.4.6. List-mode. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 431
12.5. FILE FORMAT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 431
12.5.1. File format design. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 432
12.5.2. Common image file formats. . . . . . . . . . . . . . . . . . . . . . 435
12.5.3. Movie formats. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 437
12.5.4. Nuclear medicine data requirements. . . . . . . . . . . . . . . . 437
12.5.5. Common nuclear medicine data storage formats . . . . . . 442
12.6. INFORMATION SYSTEM. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 443
12.6.1. Database . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 443
12.6.2. Hospital information system. . . . . . . . . . . . . . . . . . . . . . 445
12.6.3. Radiology information system . . . . . . . . . . . . . . . . . . . . 445
12.6.4. Picture archiving and communication system. . . . . . . . . 446
12.6.5. Scheduling. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 447
12.6.6. Broker . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 447
12.6.7. Security. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 447
CHAPTER 13. IMAGE RECONSTRUCTION. . . . . . . . . . . . . . . . . . . . . . . 449
13.1. INTRODUCTION. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 449
13.2. ANALYTICAL RECONSTRUCTION. . . . . . . . . . . . . . . . . . . . 450
13.2.1. Two dimensional tomography. . . . . . . . . . . . . . . . . . . . . 451
13.2.2. Frequencydistance relation. . . . . . . . . . . . . . . . . . . . . . 456
13.2.3. Fully 3D tomography. . . . . . . . . . . . . . . . . . . . . . . . . . . 457
13.2.4. Time of flight PET. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 466
13.3. ITERATIVE RECONSTRUCTION. . . . . . . . . . . . . . . . . . . . . . . 468
13.3.1. Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 468
13.3.2. Optimization algorithms. . . . . . . . . . . . . . . . . . . . . . . . . 473
13.3.3. Maximum-likelihood expectation-maximization . . . . . . 479
13.3.4. Acceleration. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 485
13.3.5. Regularization. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 488
13.3.6. Corrections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 495
13.4. NOISE ESTIMATION. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 507
13.4.1. Noise propagation in filtered back projection. . . . . . . . . 507
13.4.2. Noise propagation in maximum-likelihood
expectation-maximization. . . . . . . . . . . . . . . . . . . . . . . . 508

CHAPTER 14. NUCLEAR MEDICINE IMAGE DISPLAY. . . . . . . . . . . . . 512

14.1. INTRODUCTION. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 512
14.2.1. Display resolution. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 514
14.2.2. Contrast resolution. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 515
14.3. DISPLAY DEVICE HARDWARE . . . . . . . . . . . . . . . . . . . . . . . 516
14.3.1. Display controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 516
14.3.2. Cathode ray tube. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 517
14.3.3. Liquid crystal display panel. . . . . . . . . . . . . . . . . . . . . . 519
14.3.4. Hard copy devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 521
14.4. GREY SCALE DISPLAY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 521
14.4.1. Grey scale standard display function. . . . . . . . . . . . . . . . 522
14.5. COLOUR DISPLAY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 525
14.5.1. Colour and colour gamut. . . . . . . . . . . . . . . . . . . . . . . . . 528
14.6. IMAGE DISPLAY MANIPULATION . . . . . . . . . . . . . . . . . . . . 530
14.6.1. Histograms. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 530
14.6.2. Windowing and thresholding. . . . . . . . . . . . . . . . . . . . . . 530
14.6.3. Histogram equalization. . . . . . . . . . . . . . . . . . . . . . . . . . 532
14.7. VISUALIZATION OF VOLUME DATA . . . . . . . . . . . . . . . . . . 533
14.7.1. Slice mode. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 533
14.7.2. Volume mode. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 534
14.7.3. Polar plots of myocardial perfusion imaging . . . . . . . . . 538
14.8. DUAL MODALITY DISPLAY. . . . . . . . . . . . . . . . . . . . . . . . . . 540
14.9. DISPLAY MONITOR QUALITY ASSURANCE. . . . . . . . . . . . 541
14.9.1. Acceptance testing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 542
14.9.2. Routine quality control . . . . . . . . . . . . . . . . . . . . . . . . . . 542
15.1.1. Methods for routine quality assurance procedures. . . . . 547
15.2. HARDWARE (PHYSICAL) PHANTOMS. . . . . . . . . . . . . . . . . 550
15.2.1. Gamma camera phantoms. . . . . . . . . . . . . . . . . . . . . . . . 550
15.2.2. SPECT phantoms. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 558
15.2.3. PET phantoms. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 568
15.3. COMPUTATIONAL MODELS. . . . . . . . . . . . . . . . . . . . . . . . . . 575
15.3.1. Emission tomography simulation toolkits. . . . . . . . . . . . 577

15.4. ACCEPTANCE TESTING. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 578

15.4.1. Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 578
15.4.2. Procurement and pre-purchase evaluations. . . . . . . . . . . 580
15.4.3. Acceptance testing as a baseline for regular quality
assurance. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 583
15.4.4. What to do if the instrument fails acceptance testing. . . 584
15.4.5. Meeting the manufacturers specifications. . . . . . . . . . . 584
MEDICINE. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 587
16.1. INTRODUCTION. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 587
16.2. NON-IMAGING MEASUREMENTS. . . . . . . . . . . . . . . . . . . . . 588
16.2.1. Renal function measurements. . . . . . . . . . . . . . . . . . . . . 588
16.2.2. 14C breath tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 591
16.3. IMAGING MEASUREMENTS. . . . . . . . . . . . . . . . . . . . . . . . . . 591
16.3.1. Thyroid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 592
16.3.2. Renal function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 594
16.3.3. Lung function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 596
16.3.4. Gastric function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 596
16.3.5. Cardiac function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 599
CHAPTER 17. QUANTITATIVE NUCLEAR MEDICINE. . . . . . . . . . . . . . 608
MEASUREMENTS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 608
17.2.1. Region of interest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 609
17.2.2. Use of standard . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 610
17.2.3. Partial volume effect and the recovery coefficient. . . . . 610
17.2.4. Quantitative assessment . . . . . . . . . . . . . . . . . . . . . . . . . 612
17.2.5. Estimation of activity . . . . . . . . . . . . . . . . . . . . . . . . . . . 616
17.2.6. Evaluation of image quality. . . . . . . . . . . . . . . . . . . . . . 618
CHAPTER 18. INTERNAL DOSIMETRY. . . . . . . . . . . . . . . . . . . . . . . . . . . 621
FORMALISM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 621
18.1.1. Basic concepts. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 621
18.1.2. The time-integrated activity in the source region. . . . . . 626

18.1.3. Absorbed dose rate per unit activity (S value). . . . . . . . 628

18.1.4. Strengths and limitations inherent in the formalism. . . . 631
18.2.1. Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 635
18.2.2. Dosimetry on an organ level. . . . . . . . . . . . . . . . . . . . . . 636
18.2.3. Dosimetry on a voxel level. . . . . . . . . . . . . . . . . . . . . . . 637
CHAPTER 19. RADIONUCLIDE THERAPY. . . . . . . . . . . . . . . . . . . . . . . . 641
19.1. INTRODUCTION. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 641
19.2. THYROID THERAPIES. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 642
19.2.1. Benign thyroid disease . . . . . . . . . . . . . . . . . . . . . . . . . . 642
19.2.2. Thyroid cancer. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 643
19.3. PALLIATION OF BONE PAIN. . . . . . . . . . . . . . . . . . . . . . . . . . 645
19.3.1. Treatment specific issues. . . . . . . . . . . . . . . . . . . . . . . . . 646
19.4. HEPATIC CANCER. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 646
19.4.1. Treatment specific issues. . . . . . . . . . . . . . . . . . . . . . . . . 647
19.5. NEUROENDOCRINE TUMOURS. . . . . . . . . . . . . . . . . . . . . . . 647
19.5.1. Treatment specific issues. . . . . . . . . . . . . . . . . . . . . . . . . 648
19.6. NON-HODGKINS LYMPHOMA. . . . . . . . . . . . . . . . . . . . . . . 649
19.6.1. Treatment specific issues. . . . . . . . . . . . . . . . . . . . . . . . . 649
19.7. PAEDIATRIC MALIGNANCIES. . . . . . . . . . . . . . . . . . . . . . . . 650
19.7.1. Thyroid cancer. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 651
19.7.2. Neuroblastoma. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 651
19.8. ROLE OF THE PHYSICIST. . . . . . . . . . . . . . . . . . . . . . . . . . . . 652
19.9. EMERGING TECHNOLOGY. . . . . . . . . . . . . . . . . . . . . . . . . . . 654
19.10. CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 656
CHAPTER 20. MANAGEMENT OF THERAPY PATIENTS. . . . . . . . . . . . 658
20.1. INTRODUCTION. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 658
20.2. OCCUPATIONAL EXPOSURE . . . . . . . . . . . . . . . . . . . . . . . . . 658
20.2.1. Protective equipment and tools . . . . . . . . . . . . . . . . . . . 658
20.2.2. Individual monitoring . . . . . . . . . . . . . . . . . . . . . . . . . . . 659
20.3. RELEASE OF THE PATIENT. . . . . . . . . . . . . . . . . . . . . . . . . . . 659
20.3.1. The decision to release the patient. . . . . . . . . . . . . . . . . 660
20.3.2. Specific instructions for releasing the radioactive
patient . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 662
20.4. PUBLIC EXPOSURE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 665
20.4.1. Visitors to patients. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 665
20.4.2. Radioactive waste. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 665


AND WARDS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 666
20.5.1. Shielding for control of external dose. . . . . . . . . . . . . . . 666
20.5.2. Designing for control of contamination . . . . . . . . . . . . . 668
20.6. OPERATING PROCEDURES. . . . . . . . . . . . . . . . . . . . . . . . . . . 668
20.6.1. Transport of therapy doses . . . . . . . . . . . . . . . . . . . . . . . 669
20.6.2. Administration of therapeutic radiopharmaceuticals. . . 669
20.6.3. Error prevention. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 670
20.6.4. Exposure rates and postings . . . . . . . . . . . . . . . . . . . . . . 670
20.6.5. Patient care in the treating facility. . . . . . . . . . . . . . . . . . 672
20.6.6. Contamination control procedures . . . . . . . . . . . . . . . . . 673
20.7. CHANGES IN MEDICAL STATUS. . . . . . . . . . . . . . . . . . . . . . 674
20.7.1. Emergency medical procedures. . . . . . . . . . . . . . . . . . . . 675
20.7.2. The radioactive patient in the operating theatre . . . . . . . 675
20.7.3. Radioactive patients on dialysis . . . . . . . . . . . . . . . . . . . 676
20.7.4. Re-admission of patients to the treating institution. . . . . 676
20.7.5. Transfer to another health care facility. . . . . . . . . . . . . . 677
20.8. DEATH OF THE PATIENT. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 677
20.8.1. Death of the patient following radionuclide therapy. . . . 678
20.8.2. Organ donation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 679
20.8.3. Precautions during autopsy. . . . . . . . . . . . . . . . . . . . . . . 679
20.8.4. Preparation for burial and visitation . . . . . . . . . . . . . . . . 680
20.8.5. Cremation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 681
ABBREVIATIONS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 723
SYMBOLS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 729
CONTRIBUTORS to drafting and review. . . . . . . . . . . . . . . . . . . . 735

Department of Medical Physics,
McGill University,
Montreal, Canada
Division of Human Health,
International Atomic Energy Agency,
Medical Cyclotron Facility,
Board of Radiation and Isotope Technology,
Bhabha Atomic Research Centre,
Mumbai, India
The technologies used in nuclear medicine for diagnostic imaging have
evolved over the last century, starting with Rntgens discovery of X rays
and Becquerels discovery of natural radioactivity. Each decade has brought
innovation in the form of new equipment, techniques, radiopharmaceuticals,
advances in radionuclide production and, ultimately, better patient care. All
such technologies have been developed and can only be practised safely with
a clear understanding of the behaviour and principles of radiation sources and
radiation detection. These central concepts of basic radiation physics and nuclear
physics are described in this chapter and should provide the requisite knowledge
for a more in depth understanding of the modern nuclear medicine technology
discussed in subsequent chapters.
1.1.1. Fundamental physical constants
The chapter begins with a short list of physical constants of importance to
general physics as well as to nuclear and radiation physics. The data listed below
were taken from the CODATA set of values issued in 2006 and are available


from a web site supported by the National Institute of Science and Technology in
Washington, DC, United States of America: http://physics.nist.gov/cuu/Constants
Avogadros number: NA=6.022 1023 mol1 or 6.022 1023 atoms/mol.
Speed of light in vacuum: c=2.998 108 m/s 3 108 m/s.
Electron charge: e=1.602 1019 C.
Electron and positron rest mass: me=0.511 MeV/c2.
Proton rest mass: mp=938.3 MeV/c2.
Neutron rest mass: mn=939.6 MeV/c2.
Atomic mass unit: u=931.5 MeV/c2.
Plancks constant: h=6.626 1034 J s.
Electric constant (permittivity of vacuum): 0=8.854 1012 C V1 m1.
Magnetic constant (permeability of vacuum): 0=4 107 V s A1 m1.
Newtonian gravitation constant: G=6.672 1011 m3 kg1 s2.
Proton mass/electron mass: mp/me=1836.0.
Specific charge of electron: e/me=1.758 1011 C/kg.
1.1.2. Physical quantities and units
A physical quantity is defined as a quantity that can be used in mathematical
equations of science and technology. It is characterized by its numerical value
(magnitude) and associated unit. The following rules apply to physical quantities
and their units in general:
Symbols for physical quantities are set in italics (sloping type), while
symbols for units are set in roman (upright) type (e.g. m=21 kg;
E=15 MeV; K=220 Gy).
Superscripts and subscripts used with physical quantities are set in italics if
they represent variables, quantities or running numbers; they are in roman
type if they are descriptive (e.g. Nx, m but max, Eab, tr).
Symbols for vector quantities are set in bold italics.
The currently used metric system of units is known as the International
System of Units (SI). The system is founded on base units for seven basic
physical quantities. All other quantities and units are derived from the seven base
quantities and units. The seven base SI quantities and their units are:

Length l: metre (m).

Mass m: kilogram (kg).
Time t: second (s).
Electric current I: ampere (A).


(e) Temperature T: kelvin (K).

(f) Amount of substance: mole (mol).
(g) Luminous intensity: candela (cd).
Examples of basic and derived physical quantities and their units are given
in Table 1.1.

Units commonly
used in radiation

Symbol SI unit



nm, , fm

1 m=109 nm=1010 =1015 fm




1 MeV/c2=1.78 1030 kg


ms, s, ns, ps

1 s=103 ms=106 s=109 ns=1012 ps


mA, A, nA, pA

1 A=103 mA=106 A=109 nA


Mass density


Current density








Electric charge








T (in K)=T (in C) + 273.16


1 kg/m3=103 g/cm3

1 Hz=1 s1

1 e=1.602 1019 C
1 N=1 kg m s2

760 torr=101.3 kPa 1 Pa=1 N/m2 =7.5 103 torr

1 N s=1 kg m s1
eV, keV, MeV

1 eV=1.602 1019 J =103 keV

1 W=1 J/s=1 V A


1.1.3. Classification of radiation

Radiation, the transport of energy by electromagnetic waves or atomic
particles, can be classified into two main categories depending on its ability to
ionize matter. The ionization potential of atoms, i.e. the minimum energy required
to ionize an atom, ranges from a few electronvolts for alkali elements to 24.6 eV
for helium which is in the group of noble gases. Ionization potentials for all other
atoms are between the two extremes.
Non-ionizing radiation cannot ionize matter because its energy per quantum
is below the ionization potential of atoms. Near ultraviolet radiation, visible
light, infrared photons, microwaves and radio waves are examples of
non-ionizing radiation.
Ionizing radiation can ionize matter either directly or indirectly because its
quantum energy exceeds the ionization potential of atoms. X rays, rays,
energetic neutrons, electrons, protons and heavier particles are examples of
ionizing radiation.
1.1.4. Classification of ionizing radiation
Ionizing radiation is radiation that carries enough energy per quantum to
remove an electron from an atom or a molecule, thus introducing a reactive and
potentially damaging ion into the environment of the irradiated medium. Ionizing
radiation can be categorized into two types: (i) directly ionizing radiation and
(ii) indirectly ionizing radiation. Both directly and indirectly ionizing radiation
can traverse human tissue, thereby enabling the use of ionizing radiation in
medicine for both imaging and therapeutic procedures.
Directly ionizing radiation consists of charged particles, such as electrons,
protons, particles and heavy ions. It deposits energy in the medium
through direct Coulomb interactions between the charged particle and
orbital electrons of atoms in the absorber.
Indirectly ionizing radiation consists of uncharged (neutral) particles
which deposit energy in the absorber through a two-step process. In the
first step, the neutral particle releases or produces a charged particle in
the absorber which, in the second step, deposits at least part of its kinetic
energy in the absorber through Coulomb interactions with orbital electrons
of the absorber in the manner discussed above for directly ionizing charged


1.1.5. Classification of indirectly ionizing photon radiation

Indirectly ionizing photon radiation consists of three main categories:
(i) ultraviolet, (ii) X ray and (iii) ray. Ultraviolet photons are of limited use in
medicine. Radiation used in imaging and/or treatment of disease consists mostly
of photons of higher energy, such as X rays and rays. The commonly accepted
difference between the two is based on the radiations origin. The term ray
is reserved for photon radiation that is emitted by the nucleus or from other
particle decays. The term X ray, on the other hand, refers to radiation emitted
by electrons, either orbital electrons or accelerated electrons (e.g. bremsstrahlung
type radiation).
With regard to their origin, the photons of the indirectly ionizing radiation
type fall into four categories: characteristic (fluorescence) X rays, bremsstrahlung
X rays, photons resulting from nuclear transitions and annihilation quanta.
1.1.6. Characteristic X rays
Orbital electrons have a natural tendency to configure themselves in such
a manner that they inhabit a minimal energy state for the atom. When a vacancy
is opened within an inner shell, as a result of an ionization or excitation process,
an outer shell electron will make a transition to fill the vacancy, usually within
a nanosecond for solid materials. The energy liberated in this transition may be
released in the form of a characteristic (fluorescence) photon of energy equal
to the difference between the binding energies of the initial and final vacancies.
Since different elements have different binding energies for their electronic
shells, the energy of the photon released in this process will be characteristic
of the particular atom. Rather than being emitted as a characteristic photon, the
transition energy may also be transferred to an orbital electron that is then emitted
with kinetic energy that is equal to the transition energy less the electron binding
energy. The emitted orbital electron is called an Auger electron.
1.1.7. Bremsstrahlung
The word bremsstrahlung can be translated from its original German
term as braking radiation, and is a name aptly assigned to the phenomenon.
When light charged particles (electrons and positrons) are slowed down or
negatively accelerated (decelerated) by interactions with other charged particles
in matter (e.g. by atomic nuclei), the kinetic energy that they lose is converted
to electromagnetic radiation, referred to as bremsstrahlung radiation. The energy
spectrum of bremsstrahlung is non-discrete (i.e. continuous) and ranges between
zero and the kinetic energy of the initial charged particle. Bremsstrahlung plays


a central role in modern imaging and therapeutic equipment, since it can be

used to produce X rays on demand from an electrical energy source. The power
emitted in the form of bremsstrahlung photons is proportional to the square of the
particles charge and the square of the particles acceleration.
1.1.8. Gamma rays
When a nuclear reaction or spontaneous nuclear decay occurs, the process
may leave the product (daughter) nucleus in an excited state. The nucleus can
then make a transition to a more stable state by emitting a ray photon and the
process is referred to as decay. The energy of the photon emitted in decay is
characteristic of the nuclear energy transition, but the recoil of the emitting atom
produces a spectrum centred on the characteristic energy. Gamma rays typically
have energies above 100 keV and wavelengths less than 0.1 .
1.1.9. Annihilation quanta
When a parent nucleus undergoes plus decay or a high energy photon
interacts with the electric field of either the nucleus or the orbital electron, an
energetic positron may be produced. In moving through an absorber medium,
the positron loses most of its kinetic energy as a result of Coulomb interactions
with absorber atoms. These interactions result in collision loss when the
interaction is with an orbital electron of the absorber atom and in radiation loss
(bremsstrahlung) when the interaction is with the nucleus of the absorber atom.
Generally, after the positron loses all of its kinetic energy through collision and
radiation losses, it will undergo a final collision with an available orbital electron
(due to the Coulomb attractive force between the positively charged positron and
a local negatively charged electron) in a process called positron annihilation.
During annihilation, the positron and electron disappear and are replaced by
two oppositely directed annihilation quanta, each with an energy of 0.511 MeV.
This process satisfies a number of conservation laws: conservation of electric
charge, conservation of linear momentum, conservation of angular momentum
and conservation of total energy.
A percentage of positron annihilations occur before the positron expends all
of its kinetic energy and the process is then referred to as in-flight annihilation.
The two quanta emitted in in-flight annihilation are not of identical energies and
do not necessarily move in absolute opposite directions.


1.1.10. Radiation quantities and units

Accurate measurement of radiation is very important in all medical uses
of radiation, be it for diagnosis or treatment of disease. In diagnostic imaging
procedures, image quality must be optimized, so as to obtain the best possible
image with the lowest possible radiation dose to the patient to minimize the risk
of morbidity. In radiotherapy, the prescribed dose must be delivered accurately
and precisely to maximize the tumour control probability (TCP) and to minimize
the normal tissue complication probability (NTCP). In both instances, the risk
of morbidity includes acute radiation effects (radiation injury) as well as late
radiation-induced effects, such as induction of cancer and genetic damage.
Several quantities and units were introduced for the purpose of quantifying
radiation and the most important of these are listed in Table 1.2. Also listed are
the definitions for the various quantities and the relationships between the old
units and the SI units for these quantities. The definitions of radiation related
physical quantities are as follows:
Exposure X is related to the ability of photons to ionize air. Its unit, roentgen
(R), is defined as a charge of 2.58 104 coulombs produced per kilogram
of air.
Kerma K (acronym for kinetic energy released in matter) is defined for
indirectly ionizing radiation (photons and neutrons) as energy transferred to
charged particles per unit mass of the absorber.
Dose (also referred to as absorbed dose) is defined as energy absorbed per
unit mass of medium. Its SI unit, gray (Gy), is defined as 1 joule of energy
absorbed per kilogram of medium.
Equivalent dose HT is defined as the dose multiplied by a radiation
weighting factor wR. When different types of radiation are present, HT is
defined as the sum of all of the individual weighted contributions. The SI
unit of equivalent dose is the sievert (Sv).
Effective dose E of radiation is defined as the equivalent dose HT multiplied
by a tissue weighting factor wT. The SI unit of effective dose is also the
sievert (Sv).
Activity A of a radioactive substance is defined as the number of nuclear
decays per time. Its SI unit, becquerel (Bq), corresponds to one decay per




SI unit
10 4 C
kg air

Old unit

10 4 C
kg air

Exposure X X =



Kerma K


E tr

1 Gy = 1


Dose D


E ab

1 Gy = 1


dose HT


1 Sv

1 rem

1 Sv=100 rem

dose E

E = H Tw T

1 Sv

1 rem

1 Sv=100 rem

Activity A


1 Bq = 1 s 1

1 Ci = 3.7 10 10 s 1

1 Bq =


1 esu
cm 3 airSTP

1 R = 2.58

1 rad = 100


1 Gy=100 rad

1 Ci
3.7 10 10


The constituent particles forming an atom are protons, neutrons and
electrons. Protons and neutrons are known as nucleons and form the nucleus of
the atom. Protons have a positive charge, neutrons are neutral and electrons have
a negative charge mirroring that of a proton. In comparison to electrons, protons
and neutrons have a relatively large mass exceeding the electron mass by a factor
of almost 2000 (note: mp/me=1836).
The following general definitions apply to atomic structure:
Atomic number Z is the number of protons and number of electrons in an
Atomic mass number A is the number of nucleons in an atom, i.e. the
number of protons Z plus the number of neutrons N in an atom: A=Z + N.
Atomic mass ma is the mass of a specific isotope expressed in atomic mass
units u, where 1 u is equal to one twelfth of the mass of the 12C atom (unbound,
at rest and in the ground state) or 931.5 MeV/c2. The atomic mass is smaller
than the sum of the individual masses of the constituent particles because of
the intrinsic energy associated with binding the particles (nucleons) within
the nucleus. On the other hand, the atomic mass is larger than the nuclear
mass M because the atomic mass includes the mass contribution of Z orbital


electrons while the nuclear mass M does not. The binding energy of orbital
electrons to the nucleus is ignored in the definition of the atomic mass.
While for 12C the atomic mass is exactly 12 u, for all other atoms ma does
not exactly match the atomic mass number A. However, for all atomic entities,
A (an integer) and ma are very similar to one another and often the same symbol
(A) is used for the designation of both. The mass in grams equal to the average
atomic mass of a chemical element is referred to as the mole (mol) of the
element and contains exactly 6.022 1023 atoms. This number is referred to as
the Avogadro constant NA of entities per mole. The atomic mass number of all
elements is, thus, defined such that A grams of every element contain exactly
NA atoms. For example, the atomic mass of natural cobalt is 58.9332 u. Thus,
one mole of natural cobalt has a mass of 58.9332 g and by definition contains
6.022 1023 entities (cobalt atoms) per mole of cobalt.
The number of atoms Na per mass of an element is given as:

The number of electrons per volume of an element is:


= Z a = Z A (1.2)

The number of electrons per mass of an element is:


= Z A (1.3)

It should be noted that Z/A 0.5 for all elements with one notable exception
of hydrogen for which Z/A=1. Actually, Z/A slowly decreases from 0.5 for low
Z elements to 0.4 for high Z elements. For example, Z/A for 4He is 0.5, for 60Co is
0.45 and for 235U is 0.39.
If it is assumed that the mass of a molecule is equal to the sum of the masses
of the atoms that make up the molecule, then, for any molecular compound, there
are NA molecules per mole of the compound where the mole in grams is defined
as the sum of the atomic mass numbers of the atoms making up the molecule. For
example, 1 mole of water (H2O) is 18 g of water and 1 mole of carbon dioxide
(CO2) is 44 g of carbon dioxide. Thus, 18 g of water or 44 g of carbon dioxide
contain exactly NA molecules (or 3 NA atoms, since each molecule of water and
carbon dioxide contains three atoms).


1.2.1. Rutherford model of the atom

At the beginning of the 20th century, the structure of the atom was not well
known. Scientific pioneers such as Dalton, Mendeleev and Thomson, among
others, were developing a common theory through their endeavours. Often noted
as a significant contribution to the modern understanding of the atom, is the
work performed by Rutherford and his colleagues Geiger and Marsden in 1909.
Through observation of the behaviour of positively charged particles traversing
a thin gold foil, Rutherford concluded that the positive charge and most of the
mass of the atom are concentrated in the atomic nucleus (diameter of a few
femtometres) and negative electrons are spread over the periphery of the atom
(diameter of a few ngstrms). This work was significant because it introduced a
new specialty of physics (nuclear physics) and demonstrated that the atom is not
simply a single particle, but instead is made up of smaller subatomic particles,
organized in an atom with well defined characteristics.
1.2.2. Bohr model of the hydrogen atom
Bohr expanded the Rutherford atomic model in 1913 using a set of four
postulates that combine classical, non-relativistic mechanics with the concept
of angular momentum quantization. The Bohr model of the atom can be said to
resemble a planetary model in that the protons and neutrons occupy a dense
central region called the nucleus and the electrons orbit the nucleus as planets
orbit the sun. The Bohr model introduces the concept that the angular momenta
of orbital electrons revolving around the nucleus in allowed orbits, radii of the
allowed electronic orbits (shells), velocities of orbital electrons in allowed orbits
and binding energies of orbital electrons in allowed orbits within the atom, are
restricted to certain discrete states. This means that angular momenta, radii,
velocities and binding energies of orbital electrons are quantized.
While scientific theory was later expanded to include the necessary
principles of quantum mechanics in our understanding of the atom, the Bohr
model is elegant and provides a simplistic, yet practical, view of the atom that is
still used for teaching atomic principles, and successfully deals with one-electron
entities, such as the hydrogen atom, the singly ionized helium atom and the
doubly ionized lithium atom.
According to the RutherfordBohr atomic model, most of the atomic mass
is concentrated in the atomic nucleus consisting of Z protons and (A Z) neutrons,


where Z is the atomic number and A the atomic mass number of a given nucleus.
In nuclear physics, the convention is to designate a nucleus X as ZA X , where A is
its atomic mass number and Z its atomic number; for example, the 60Co nucleus
is identified as 60
Ra nucleus as 226
27 Co and the
88 Ra. The atomic number Z is
often omitted in references to an atom because the atom is already identified by
its 13 letter symbol. In ion physics, the convention is to designate ions with +
or superscripts. For example, 42 He + stands for a singly ionized helium atom
and 42 He 2+ stands for a doubly ionized helium atom, also known as the particle.
With regard to relative values of atomic number Z and atomic mass number A of
nuclei, the following conventions apply:
An element may be composed of atoms that all have the same number of
protons, i.e. have the same atomic number Z, but have a different number of
neutrons (have different atomic mass numbers A). Such atoms of identical
Z but differing A are called isotopes of a given element.
The term isotope is often misused to designate nuclear species. For
example, 60Co, 137Cs and 226Ra are not isotopes, since they do not belong
to the same element. Rather than isotopes, they should be referred to as
nuclides. On the other hand, it is correct to state that deuterium (with a
nucleus called deuteron) and tritium (with a nucleus called triton) are heavy
isotopes of hydrogen or that 59Co and 60Co are isotopes of cobalt. Thus,
the term radionuclide should be used to designate radioactive species;
however, the term radioisotope is often used for this purpose.
A nuclide is an atomic species characterized by its nuclear composition (A, Z
and the arrangement of nucleons within the nucleus). The term nuclide
refers to all atomic forms of all elements. The term isotope is narrower
and only refers to various atomic forms of a given chemical element.
In addition to being classified into isotopic groups (common atomic
number Z), nuclides are also classified into groups with a common atomic mass
number A (isobars) and a common number of neutrons (isotones). For example,
Co and 60Ni are isobars with 60 nucleons each (A=60), and 67
31 Ga , 32 Ge and
33 As are isobars with atomic mass number 67, while H (tritium) and He are
isotones with two neutrons each (A Z=2), and 126 C, 137 N and 148 O are isotones
with six neutrons each.
A tool for remembering these definitions is as follows: isotopes have the
same number of protons Z; isotones have the same number of neutrons, A Z;
isobars have the same mass number A.
If a nucleus exists in an excited state for some time, it is said to be in
an isomeric (metastable) state. Isomers are, thus, nuclear species that have a



common atomic number Z and a common atomic mass number A. For example,
Tc is an isomeric state of 99Tc and 60mCo is an isomeric state of 60Co.
1.3.1. Nuclear radius
The radius R of a nucleus with atomic mass number A is estimated from the
following expression:
R = R0 3 A


where R0 is the nuclear radius constant equal to 1.25 fm. Since the range of A in
nature is from 1 to about 250, nuclear radius ranges from about 1 fm for a proton
to about 8 fm for heavy nuclei.
1.3.2. Nuclear binding energy
The sum of the masses of the individual components of a nucleus that
contains Z protons and (A Z) neutrons is larger than the actual mass of the
nucleus. This difference in mass is called the mass defect (deficit) m and its
energy equivalent mc2 is called the total binding energy EB of the nucleus. The
total binding energy EB of a nucleus can, thus, be defined as the energy liberated
when Z protons and (A Z) neutrons are brought together to form the nucleus.
The binding energy per nucleon (EB/A) in a nucleus (i.e. the total binding
energy of a nucleus divided by the number of nucleons in the given nucleus)
varies with the number of nucleons A and is of the order of ~8 MeV/nucleon.
A plot of the binding energy per nucleon EB/A in megaelectronvolts per
nucleon against the atomic mass number in the range from 1 to 250 is given in
Fig.1.1 and shows a rapid rise in EB/A at small atomic mass numbers, a broad
maximum of about 8.7 MeV/nucleon around A 60 and a gradual decrease in
EB/A at large A. The larger the binding energy per nucleon (EB/A) of an atom,
the larger is the stability of the atom. Thus, the most stable nuclei in nature are
the ones with A 60 (iron, cobalt, nickel). Nuclei of light elements (small A) are
generally less stable than nuclei with A 60, and the heaviest nuclei (large A) are
also less stable than nuclei with A 60.


(small A) are generally less stable than nuclei with A 60, and the heaviest nuclei (large A)
are also less stable than nuclei with A 60.

FIG.1.1. Binding energy per nucleon in megaelectronvolts per nucleon against atomic mass

FIG. 1.1.
Binding energy per nucleon in megaelectronvolts per nucleon against atomic mass
number A. Data are from the National Institute of Science and Technology (NIST).
number A. Data are from the National Institute of Science and Technology (NIST).
1.3.3. Nuclear fusion and fission
1.3.3. Nuclear fusion and fission
The peculiar
EB/A versus
A (Fig.
curve 1.1)
two for
The peculiar
shape shape
of theofEBthe
/A versus
A curve
two methods
converting mass into energy: (i) fusion of nuclei at low A and (ii) fission of nuclei at and
large A:
(ii) fission of nuclei at large A:
Fusion of two nuclei of very small mass, e.g. 21 H + 31 H 24 He + n , will
create a more massive nucleus and release a certain amount of energy.
Experiments using controlled nuclear fusion for production of energy have
so far not been successful in generating a net energy gain, i.e. the amount
of energy consumed is still larger than the amount created. However, fusion
remains an active field of research and it is reasonable to expect that in the
future controlled fusion will play an important role in the production of
electrical power.
Fission attained by bombardment of certain elements of large mass (such as
U) by thermal neutrons in a nuclear reactor will create two lower mass
and more stable nuclei, and transform some mass into kinetic energy of the
two product nuclei. Hahn, Strassman, Meitner and Frisch described fission
in 1939, and, in 1942, Fermi and colleagues at the University of Chicago
carried out the first controlled chain reaction based on nuclear fission.


Since then, fission reactors have become an important means of production

of electrical power.
1.3.4. Two-particle collisions and nuclear reactions
A common problem in nuclear physics and radiation dosimetry is the
collision of two particles in which a projectile with mass m1, velocity 1 and
kinetic energy (EK)1 strikes a stationary target with mass m2 and velocity 2=0.
The probability or cross-section for a particular collision as well as the collision
outcome depends on the physical properties of the projectile (mass, charge,
velocity, kinetic energy) and the stationary target (mass, charge).
As shown schematically in Fig.1.2, the collision between the projectile
and the target in the most general case results in an intermediate compound that
subsequently decays into two reaction products: one of mass m3 ejected with
velocity 3 at an angle to the incident projectile direction, and the other of
mass m4 ejected with velocity 4 at an angle to the incident projectile direction.
Two-particle collisions are classified into three categories: (a) elastic
scattering, (b) inelastic collisions and (c) nuclear reactions:
(a) Elastic scattering is a special case of a two-particle collision in which the
products after the collision are identical to the products before collision,
i.e. m3=m1 and m4=m2, and the total kinetic energy and momentum
before the collision are equal to the total kinetic energy and momentum,
respectively, after the collision.
(b) In inelastic scattering of a projectile m1 on the target m2, similarly to elastic
scattering, the reaction products after collision are identical to the initial
products, i.e. m3=m1 and m4=m2; however, the incident projectile transfers
a portion of its kinetic energy to the target in the form of not only kinetic
energy but also intrinsic excitation energy E*.
(c) During a nuclear reaction, a collision between a projectile m1 and a target
m2 takes place and will result in the formation of two reaction products
m3 and m4, with the products having new atomic numbers. This process
is shown schematically in Fig.1.2. In any nuclear reaction, a number
of physical quantities must be conserved, most notably charge, linear
momentum and massenergy. In addition, the sum of atomic numbers Z
and the sum of atomic mass numbers A before and after the collision must
also be conserved.



The Q value of a nuclear reaction is defined as the difference between the

total rest energy before the reaction (m1c 2 + m 2c 2 ) and the total rest energy
after the reaction (m 3c 2 + m 4c 2 ) or:
Q = (m1c 2 + m 2c 2 ) (m 3c 2 + m 4c 2 ) (1.5)

Each Qtwo-particle
collision possesses a characteristic Q value(1.5)
that can
= (m c + m c ) - (m c + m c )
be either positive, zero or negative. For Q > 0, the collision is termed
Each two-particle collision possesses a characteristic Q value that can be either
(also called exoergic) and results in a release of energy;
positive, zero or negative. For Q > 0, the collision is termed exothermic (also called
for Q=0,
is termed
Q <is termed
0, theelastic
collision is
results in a release
of energy;
for Q = and
0, the for
for Q < 0, the collision is termed endothermic (also called endoergic), and to
it requires
take place, it requires an energy transfer from the projectile to the target. An
an energy
to while
the target.
An exothermic
can the
an endothermic
reaction cannot
take place unless the projectile has kinetic energy exceeding the threshold energy
can occur
(EK)thr given as:
unless the projectile has kinetic energy exceeding the threshold energy
(EK)thr given as:

(m 3c 2 + m 4c 2 ) 2 - (m1c 2 + m 2c 2 ) 2


(E ) 2 =
-Q1 +
m 2
c 2 c 2 + m c 2 ) 2
(m 3Kc thr
+ m 4c 2 ) 2 2m(2m
1 + m1 (1.6)
(E K ) thr =

m 2
2m 2c 2

m1c2m,1cm2, 2mc22c,2,m
4c theare
m33cc2 and
rest the
the projectile
, target
products mproducts
target and
m2 reaction
and reaction
m3 and m4, respectively.
3 and m4, respectively.

FIG. 1.2. Schematic representation of a two-particle collision of a projectile (incident

of mass mrepresentation
a stationary target
with mass
velocity (incident
1 and velocity 1 striking
of a two-particle
of ma2 and
= 0. An intermediate compound entity is formed temporarily that subsequently decays into

mass m
1 m4.
of mass m3 and
intermediate compound entity is formed temporarily that subsequently decays into
2=0. 1.4.
two reaction products of mass m3 and m4.



Radioactivity, also known as radioactive decay, nuclear decay, nuclear
disintegration and nuclear transformation, is a spontaneous process by which
an unstable parent nucleus emits a particle or electromagnetic radiation and
transforms into a more stable daughter nucleus that may or may not be stable.
The unstable daughter nucleus will decay further in a decay series until a stable
nuclear configuration is reached. Radioactive decay is usually accompanied by
emission of energetic particles or ray photons or both.
All radioactive decay processes are governed by the same general
formalism that is based on the definition of the activity A(t) and on a characteristic
parameter for each radioactive decay process, the radioactive decay constant l
with dimensions of reciprocal time, usually in s1. The main characteristics of
radioactive decay are as follows:
The radioactive decay constant multiplied by a time interval that is much
smaller than 1/ represents the probability that any particular atom of a
radioactive substance containing a large number N(t) of identical radioactive
atoms will decay (disintegrate) in that time interval. An assumption is made
that is independent of the physical environment of a given atom.
The activity A(t) of a radioactive substance containing a large number
N(t) of identical radioactive atoms represents the total number of decays
(disintegrations) per unit time and is defined as a product between N(t) and
, i.e.:
A (t ) = N (t ) (1.7)

The SI unit of activity is the becquerel (Bq) given as 1 Bq=1 s1. The
becquerel and hertz both correspond to s1, but hertz refers to the frequency of
periodic motion, while becquerel refers to activity.
The old unit of activity, the curie (Ci), was initially defined as the activity
of 1 g of 226Ra; 1 Ci @ 3.7 1010 s1.
Subsequently, the activity of 1 g of 226Ra was determined to be
3.665 1010 s1; however, the definition of the activity unit curie (Ci) was
kept as 1 Ci=3.7 1010 s1. Since the unit of activity the becquerel is 1 s1,
the SI unit becquerel (Bq) and the old unit curie (Ci) are related as follows:
1 Ci=3.7 1010 Bq and, consequently, 1 Bq=(3.7 1010)1 Ci=2.703 1011 Ci.



Specific activity a is defined as activity A per unit mass m, i.e.:





where NA is Avogadros number.

Specific activity a of a radioactive atom depends on the decay constant
and on the atomic mass number A of the radioactive atom. The units of specific
activity are Bq/kg (SI unit) and Ci/g (old unit).
1.4.1. Decay of radioactive parent into a stable or unstable daughter
The simplest form of radioactive decay involves a radioactive parent
nucleus P decaying with decay constant P into a stable or unstable daughter
nucleus D:

D (1.9)

The rate of depletion of the number of radioactive parent nuclei NP(t) is

equal to activity AP(t) at time t defined as the product N(t) in Eq.(1.7). We, thus,
have the following expression:
d N P (t )
= A P (t ) = P N P (t ) (1.10)

The fundamental differential equation in Eq.(1.10) for NP(t) can be

rewritten in general integral form:
N P (t )

N P (0)

d N P (t )

dt (1.11)

where NP(0) is the initial condition represented by the number of radioactive

nuclei at time t=0.



Assuming that P is constant, Eq.(1.11) can be solved to obtain:


N P (t )
= P t (1.12)
N P (0)

N P (t ) = N P (0)e P t (1.13)

Based on the definition of activity given in Eq.(1.7), the activity of parent

nuclei P at time t can be expressed as follows:
A P (t ) = P N P (t ) = P N P (0)e P t = A P (0)e P t


where AP(0)=PNP(0) is the initial activity of the radioactive substance.

The decay law of Eq.(1.14) applies to all radioactive nuclides irrespective
of their mode of decay; however, the decay constant P is different for each
parent radioactive nuclide P and is the most important defining characteristic of
a radioactive nuclide.
Two special time periods called half-life (T1/2)P and mean or average life P
are used to characterize a given radioactive parent substance P. The half-life (T1/2)P
of a radioactive substance P is the time during which the number of radioactive
nuclei of the substance decays to half of the initial value NP(0) present at time
t=0. It can also be stated that in the time of one half-life the activity AP(t) of a
radioactive substance decreases to one half of its initial value AP(0)=PNP(0):
N P [t = (T1/2 ) P ] = N P (0) = N P (0)e P (T1/2 ) P (1.15)

A P [t = (T1/2 ) P ] =

A (0) = A P (0)e P (T1/2 ) P (1.16)
2 P

From Eqs(1.15) and (1.16), it is noted that e P (T1/2 ) P must equal 1/2,
resulting in the following relationship between the decay constant P and half-life
P =


ln 2
(T1/2 ) P (T1/2 ) P


Mean (average) life P of a radioactive parent P is defined as the time

required for the number NP of radioactive atoms or its activity AP to fall to
1/e=0.368 (or 36.8%) of the initial number of nuclei NP(0) or of the initial
activity AP(0), respectively. Thus, the following expressions describe the mean
N P (t = P ) = N P (0) = 0.368 N P (0) = N P (0)e P P (1.18)

A P (t = P ) = A P (0) = 0.368 A P (0) = A P (0)e P P (1.19)

From Eqs(1.18) and (1.19), it is noted that e l P P must be equal to

1/e=e1=0.368, resulting in PP=1 and P=1/P. We now get the following
relationship between mean life P and half-life (T1/2)P using Eq.(1.17) and
P =

ln 2

(T1/2 ) P P


P =

(T1/2 ) P
= 1.44(T1/2 ) P (1.21)
ln 2

A typical example of a radioactive decay for initial condition

AP(t=0)=AP(0) is shown in Fig.1.3 with a plot of parent activity AP(t) against
time t given in Eq.(1.14).
1.4.2. Radioactive series decay
The radioactive decay of parent P into stable daughter D, discussed in
Section 1.4.1, is the simplest known radioactive decay process; however, the
decay of a radioactive parent P with decay constant P into a radioactive (unstable)
daughter D which in turn decays with decay constant D into a stable or unstable
G ), is much more common and
grand-daughter G, i.e. ( P
results in a radioactive decay series for which the last decay product is stable.
The parent P in the decay series follows a straightforward radioactive
decay described by Eq.(1.16) for the rate of change of the number of parent


FIG. 1.3. Activity A (t) plotted against time t for a simple decay of a radioactive parent P into
FIG. 1.3. Activity APP(t) plotted against time t for a simple decay of a radioactive parent P into
a stable or unstable daughter D. The concepts of half-life (T1/2)P and mean life P are also
a stable or unstable daughter D. The concepts of half-life (T1/2)P and mean life P are also
illustrated. The
The area
area under
under the
the exponential
exponential decay
fromtt==0 0totot t==isisequal
product APP(0)PP where APP(0) is the initial activity of the parent P. The slope of thetangent

at Pt. =
the decay curve at t = 0 is equal to PAP(0) and this tangent crosses the abscissa axis

nuclei dNP(t)/dt. The rate of change of the number of daughter nuclei dND(t)/dt,
however, is more complicated and consists of two components, one being the
1.4.2. Radioactive series decay
supply of new daughter nuclei D through the decay of P given as PNP(t) and
the other
being the
D fromD,the
decay of
D to G 1.4.1,
into stable
in Section
however, the
of a radioactive parent P with
in thedecay

decay constant P into a radioactive (unstable) daughter D which in turn decays with decay
constant D into a stable or unstable granddaughter G, i.e. ( P
D G ), is much
dN D (t )
P t
more common
decay series
forD (which
the last decay product is
) DN
) = P N P (0)e
t ) (1.22)
= PN
P (tresults
The parent P in the decay series follows a straightforward radioactive decay described
Eq. the
for the
rate of change
of the
of parent
rate of
(i) the
for time
of the
of daughter
nuclei dN
is are
of parent
P is
of two components, one being the supply of new daughter nuclei D through the decay of P
present, i.e. ND(t=0)=0, the solution of the differential equation in Eq.(1.22)
given as PNP(t) and the other being the loss of daughter nuclei D from the decay of D to G
DND(t), resulting in the following expression for dND(t)/dt:

d N (t )

= l P N P (tP) - l D eND(Ptt)
= leP
- l D N D (t )
N D (t ) =dN
t P

-l P t

With the initial conditions for time t = 0 assuming that (i) the initial number of parent nuclei P
is NP(t = 0) = NP(0), and (ii) there are no daughter D nuclei present, i.e. ND(t = 0) = 0, the
solution of the differential equation in Eq. (1.22) reads as follows:



Recognizing that the activity of the daughter AD(t) is DND(t), the daughter
activity AD(t) is written as:
D P P t
D P t
e D t = A P (0)
e D t


1 P t
e D t = A P (t )
=A P (0)
1 e ( D P )t


A D (t ) = N P (0)


AD(t) is the activity at time t of the daughter nuclei equal to DND(t);
AP(0) is the initial activity of the parent nuclei present at time t=0;
and AP(t) is the activity at time t of the parent nuclei equal to PNP(t).
While for initial conditions AP(t=0)=AP(0) and AD(t=0)=0, the parent P
activity AP(t) follows the exponential decay law of Eq.(1.14) shown in Fig.1.3,
the daughter D activity AD(t) starts at 0, then initially rises with time t, reaches
a maximum at a characteristic time t=(tmax)D, and then diminishes to reach 0 at
t=. The characteristic time (tmax)D is given as follows:
(t max )D =

1.4.3. Equilibrium in parentdaughter activities

In many parent P daughter D grand-daughter G relationships, after a
certain time t the parent and daughter activities reach a constant ratio independent
of a further increase in time t. This condition is referred to as radioactive
equilibrium and can be analysed by examining the behaviour of the activity ratio
AD(t)/AP(t) obtained from Eq.(1.24) as:
A D (t )
1 e ( D P )t =
1 e ( D P )t (1.26)

A P (t ) D P



Three possibilities merit special consideration:

(a) The half-life of the daughter exceeds that of the parent: (T1/2)D > (T1/2)P
resulting in D < P. The activity ratio AD(t)/AP(t) of Eq.(1.26) is written as:
A D (t )
D ( P D ) t
1 (1.27)

A P (t ) P D

The ratio AD(t)/AP(t) increases exponentially with time t, indicating that no

equilibrium between the parent activity AP(t) and daughter activity AD(t)
will be reached.
(b) The half-life of the daughter is shorter than that of the parent: (T1/2)D < (T1/2)P
or D > P.
The activity ratio AD(t)/AP(t) at large t becomes a constant equal to
D/(D P) and is then independent of t and larger than unity, implying
transient equilibrium, i.e.:
A D (t )
= const > 1
A P (t ) D P


(c) The half-life of the daughter is much shorter than that of the parent:
(T1/2)D (T1/2)P or D P.
For relatively large time t tmax, the activity ratio AD(t)/AP(t) of Eq.(1.28)
simplifies to:
A D (t )
1 (1.29)
A P (t )

The activity of the daughter AD(t) very closely approximates that of its parent
AP(t), i.e. AD(t) AP(t), and they decay together at the rate of the parent. This
special case of transient equilibrium in which the daughter and parent activities
are essentially identical is called secular equilibrium.
1.4.4. Production of radionuclides (nuclear activation)
In 1896, Henri Becquerel discovered natural radioactivity, and in 1934
Frdric Joliot and Irne Curie-Joliot discovered artificial radioactivity. Most
natural radionuclides are produced through one of four radioactive decay chains,
each chain fed by a long lived and heavy parent radionuclide. The vast majority
of currently known radionuclides, however, are human-made and artificially
produced through a process of nuclear activation which uses bombardment of a


stable nuclide with a suitable energetic particle or high energy photons to induce
a nuclear transformation. Various particles or electromagnetic radiation generated
by a variety of machines are used for this purpose, most notably neutrons from
nuclear reactors for neutron activation, protons from cyclotrons or synchrotrons
for proton activation, and X rays from high energy linear accelerators for nuclear
Neutron activation is important in production of radionuclides used
for external beam radiotherapy, brachytherapy, therapeutic nuclear medicine
and nuclear medicine imaging also referred to as molecular imaging; proton
activation is important in production of positron emitters used in positron
emission tomography (PET) imaging; and nuclear photoactivation is important
from a radiation protection point of view when components of high energy
radiotherapy machines become activated during patient treatment and pose a
potential radiation risk to staff using the equipment.
A more in depth discussion of radionuclide production can be found in
Chapter 4.
1.4.5. Modes of radioactive decay
Nucleons are bound together to form the nucleus by the strong nuclear
force that, in comparison to the protonproton Coulomb repulsive force, is at
least two orders of magnitude larger but of extremely short range (only a few
femtometres). To bind the nucleons into a stable nucleus, a delicate equilibrium
between the number of protons and the number of neutrons must exist. For light
(low A) nuclear species, a stable nucleus is formed by an equal number of protons
and neutrons (Z=N). Above the nucleon number A 40, more neutrons than
protons must constitute the nucleus to form a stable configuration in order to
overcome the Coulomb repulsion among the charged protons.
If the optimal equilibrium between protons and neutrons does not exist, the
nucleus is unstable (radioactive) and decays with a specific decay constant into
a more stable configuration that may also be unstable and decay further, forming
a decay chain that eventually ends with a stable nuclide.
Radioactive nuclides, either naturally occurring or artificially produced
by nuclear activation or nuclear reactions, are unstable and strive to reach
more stable nuclear configurations through various processes of spontaneous
radioactive decay that involve transformation to a more stable nuclide and
emission of energetic particles. General aspects of spontaneous radioactive decay
may be discussed using the formalism based on the definitions of activity A and
decay constant without regard for the actual microscopic processes that underlie
the radioactive disintegrations.



A closer look at radioactive decay processes shows that they are divided
into six categories, consisting of three main categories of importance to
medical use of radionuclides and three categories of less importance. The main
categories are: (i) alpha () decay, (ii) beta () decay encompassing three related
decay processes (beta minus, beta plus and electron capture) and (iii) gamma
() decay encompassing two competing decay processes (pure decay and
internal conversion). The three less important radioactive decay categories are:
(i) spontaneous fission, (ii) proton emission decay and (iii) neutron emission
Nuclides with an excess number of neutrons are referred to as neutron-rich;
nuclides with an excess number of protons are referred to as proton-rich. The
following features are notable:
For a slight protonneutron imbalance in the nucleus, radionuclides decay
by decay characterized by transformation of a proton into a neutron in
+ decay, and transformation of a neutron into a proton in decay.
For a large protonneutron imbalance in the nucleus, the radionuclides
decay by emission of nucleons: particles in decay, protons in proton
emission decay and neutrons in neutron emission decay.
For very large atomic mass number nuclides (A > 230), spontaneous fission,
which competes with decay, is also possible.
Excited nuclei decay to their ground state through decay. Most of these
transformations occur immediately upon production of the excited state by either
or decay; however, a few exhibit delayed decays that are governed by their
own decay constants and are referred to as metastable states (e.g. 99mTc).
Nuclear transformations are usually accompanied by emission of energetic
particles (charged particles, neutral particles, photons, etc.). The particles released
in the various decay modes are as follows:
Alpha particles in decay;
Electrons in decay;
Positrons in + decay;
Neutrinos in + decay;
Antineutrinos in decay;
Gamma rays in decay;
Atomic orbital electrons in internal conversion;
Neutrons in spontaneous fission and in neutron emission decay;
Heavier nuclei in spontaneous fission;
Protons in proton emission decay.



In each nuclear transformation, a number of physical quantities must

be conserved. The most important of these quantities are: (i) total energy,
(ii) momentum, (iii) charge, (iv) atomic number and (v) atomic mass number
(number of nucleons).
The total energy of particles released by the transformation process is
equal to the net decrease in the rest energy of the neutral atom, from parent P to
daughter D. The disintegration (decay) energy, often referred to as the Q value
for the radioactive decay, is defined as follows:
Q = {M (P) [M (D) + m]}c 2 (1.30)

where M(P), M(D) and m are the nuclear rest masses (in unified atomic mass
units u) of the parent, daughter and emitted particles, respectively.
For radioactive decay to be energetically possible, the Q value must be
greater than zero. This means that spontaneous radioactive decay processes
release energy and are called exoergic or exothermic. For Q > 0, the energy
equivalent of the Q value is shared as kinetic energy between the particles emitted
in the decay process and the daughter product. Since the daughter generally has a
much larger mass than the other emitted particles, the kinetic energy acquired by
the daughter is usually negligibly small.
1.4.6. Alpha decay
In decay, a radioactive parent nucleus P decays into a more stable
daughter nucleus D by ejecting an energetic particle. Since the particle is a
He nucleus ( 42 He 2+ ), in decay the parents atomic number Z decreases by two
and its atomic mass number A decreases by four:
Z P Z 2 d + 2 he

Z 2 d + a


Naturally occurring particles have kinetic energies between 4 and 9 MeV;

their range in air is between 1 and 10 cm, and their range in tissue is between
10 and 100 m.



Typical examples of decay are the decay of 226Ra with a half-life of

1602 a into 222Rn which is also radioactive and decays with a half-life of 3.82 d
into 218Po:
88 Ra
86 Rn + a
T =1602 a

86 Rn
84 Po + a
T = 3.82 d


1.4.7. Beta minus decay

In beta minus () decay, a neutron-rich parent nucleus P transforms
a neutron into a proton and ejects an electron e and an electronic
antineutrino e . Thus, in decay, the atomic number of the daughter increases
by one, i.e. ZD=ZP + 1, the atomic mass number remains constant, i.e. AD=AP,
and the general relationship for decay is given as:

Z P Z +1 D + e + e (1.33)

A typical example of decay is the decay of 60Co with a half-life of

5.26 a into an excited state of 60Ni ( 60
28 Ni ). Excited states of Ni progress to the
ground state of Ni instantaneously (within 10 s) through emission of rays
in decay:
27 Co
28 Ni
T = 5.26 a

+ e + e (1.34)

1.4.8. Beta plus decay

In beta plus (+) decay, a proton-rich parent nucleus P transforms a proton
into a neutron and ejects a positron e+ and an electronic neutrino e . Thus, in +
decay, the atomic number of the daughter decreases by one, i.e. ZD=ZP 1, the
atomic mass number, just as in decay, remains constant, i.e. AD =AP, and the
general relationship for + decay is written as:
Z P Z 1 D + e

+ e (1.35)

Radionuclides undergoing + decay are often called positron emitters and

are used in medicine for functional imaging with the special imaging technique


PET. The most common tracer for PET studies is fluorodeoxyglucose (FDG)
labelled with 18F which serves as a good example of + decay:
9 F
T =110 min

+ e (1.36)

1.4.9. Electron capture

Electron capture radioactive decay may occur when an atomic electron
ventures inside the nuclear volume, is captured by a proton, triggers a proton to
neutron transformation, and an ejection of an electronic neutrino e. In electron
capture, as in + decay, the atomic number of the daughter decreases by one,
i.e. ZD=ZP 1, and its atomic mass number, just as in and + decay, remains
constant, i.e. AD = AP. The general relationship for electron capture decay is
written as:

Z P+e

Z 1 D + e (1.37)
T =60 d

A simple example of electron capture decay is the decay of 125I with a

half-life of 60 d into an excited state of 125Te which then decays to the ground
state of 125Te through decay and internal conversion:

53 I + e


52 Te + e
T =60 d


1.4.10. Gamma decay and internal conversion

Alpha decay as well as the three decay modes (, + and electron capture)
may produce a daughter nucleus in an excited state without expending the full
amount of the decay energy available. The daughter nucleus will then reach its
ground state, either instantaneously or with some time delay (isomeric metastable
state), through one of the following two processes:
(a) By emitting the excitation energy in the form of one or more photons in a
decay process referred to as decay.
(b) By transferring the excitation energy to one of its associated atomic
orbital electrons (usually a K shell electron) in a process called internal
conversion. The vacancy left behind by the ejected orbital electron is filled
by a transition from a higher atomic shell, resulting in characteristic X rays
and/or Auger electrons.



In most radioactive or decays, the daughter nucleus de-excitation

occurs instantaneously (i.e. within 1012 s), so that the emitted rays are referred
to as if they were produced by the parent nucleus (e.g. 60Co rays). The rays
produced from isomeric transitions are attributed to the isomeric daughter
product (e.g. 99mTc rays).
The decay process and the internal conversion process may be represented,
respectively, as follows:
A *


A +


A *

+ e



A *

stands for an excited state of the nucleus

and ZA X + is the singly ionized state of atom


Z X;

following internal conversion

An example of decay is the transition of an excited 60
28 Ni nucleus,

resulting from the decay of 27 Co , into stable 28 Ni through an emission of two
rays with energies of 1.17 and 1.33 MeV. An example of internal conversion
decay is the decay of excited 125
52 Te which results from an electron capture decay
of 125I into stable 125
52 Te through emission of 35 keV rays (7 %) and internal
conversion electrons (93%).

1.4.11. Characteristic (fluorescence) X rays and Augerelectrons

A large number of radionuclides used in nuclear medicine (e.g. 99mTc, 123I,
Tl, 64Cu) decay by electron capture and/or internal conversion. Both processes
leave the atom with a vacancy in an inner atomic shell, most commonly the
K shell. The vacancy in the inner shell is filled by an electron from a higher level
atomic shell and the binding energy difference between the two shells is either
emitted as a characteristic X ray (fluorescence photon) or transferred to a higher
shell orbital electron which is then emitted from the atom as an Auger electron
with a kinetic energy equal to the transferred energy minus the binding energy
of the emitted Auger electron. Emission of characteristic photons and Auger
electrons is discussed further in Section 1.6.4.




As an energetic charged particle, such as an electron or positron, traverses
an absorbing medium, it experiences a large number of Coulomb interactions
with the nuclei and orbital electrons of absorber atoms before its kinetic energy is
expended. In each interaction, the charged particles path may be altered (elastic
or inelastic scattering) and it may lose some of its kinetic energy that will be
transferred to the medium or to photons. Charged particle interactions with orbital
electrons of the absorber result in collision (ionization loss); interactions with
nuclei of the absorber result in radiation loss. Each of these possible interactions
between the charged particle and absorber atom is characterized by a specific
cross-section (probability) for the particular interaction. The energy loss of the
charged particle propagating through an absorber depends on the properties of the
particle, such as its mass, charge, velocity and energy, as well as on the properties
of the absorber, such as its density and atomic number.
Stopping power is a parameter used to describe the gradual loss of energy
of the charged particle as it penetrates into an absorbing medium. Two classes
of stopping power are known: collision stopping power scol results from charged
particle interaction with orbital electrons of the absorber and radiation stopping
power srad results from charged particle interaction with nuclei of the absorber.
The total stopping power stot is the sum of the collision stopping power and the
radiation stopping power.
1.5.1. Electronorbital interactions
Coulomb interactions between the incident electron or positron and orbital
electrons of an absorber result in ionizations and excitations of absorber atoms.
Ionization is described as ejection of an orbital electron from the absorber atom,
thereby producing an ion. Excitation, on the other hand, is defined as the transfer
of an orbital electron of the absorber atom from an allowed orbit to a higher
allowed orbit (shell), thereby producing an excited atom. Atomic excitations and
ionizations result in collision energy loss and are characterized by collision (also
known as ionization) stopping powers.
1.5.2. Electronnucleus interactions
Coulomb interactions between the incident electron or positron and nuclei
of the absorber atom result in particle scattering. The majority of these scattering
events are elastic and result in no energy loss. However, when the scattering is
inelastic, the incident charged particle loses part of its kinetic energy through
production of X ray photons referred to as bremsstrahlung radiation. This energy


loss is characterized by radiation stopping powers and is governed by the Larmor

relationship which states that the rate of energy loss is proportional to the square
of the particles acceleration and the square of the particles charge.
1.6.1. Exponential absorption of photon beam in absorber
The most important parameter used in characterization of X ray or ray
penetration into absorbing media is the linear attenuation coefficient . This
coefficient depends on the energy h of the photon and the atomic number Z of
the absorber, and may be described as the probability per unit path length that
a photon will have an interaction with the absorber. The attenuation coefficient
is determined experimentally by aiming a narrowly collimated monoenergetic
photon beam h onto a suitable radiation detector and placing an absorber
material of varying thickness x between the photon source and the detector. The
absorber decreases the detector signal intensity from I(x) which is measured with
no absorber in the beam (x=0) to I(x) measured with an absorber of thickness
x > 0 in the beam.
dI(x)/dx, the rate of change in beam intensity I(x) transmitted through an
absorber of thickness x, is equal to the product of the attenuation coefficient
and the beam intensity I(x) at thickness x (seeEq.(1.41)). Alternatively, it can
be said that an absorber of thickness dx reduces the beam intensity by dI and the
fractional reduction in intensity dI/I is equal to the product of the attenuation
coefficient and the absorber layer thickness dx (seeEq.(1.42)). The following
expressions are obtained, respectively:
dI (x)
= I (x) (1.41)


= dx (1.42)

where the negative sign is used to indicate a decrease in signal I(x) with an
increase in absorber thickness x.
It should be noted that Eqs(1.41) and (1.42) can be considered identical.



The form of Eq.(1.41) is identical to the form of Eq.(1.10) that deals with
simple radioactive decay; however, it must be noted that in radioactive decay the
product N(t) is defined as activity A(t), while in photon beam attenuation the
product I(x) does not have a special name and symbol.
Integration of Eq.(1.42) over absorber thickness x from 0 to x and over
intensity I(x) from the initial intensity I(0) (no absorber) to intensity I(x) at
absorber thickness x, gives:
I ( x)

I (0)


dx (1.43)

resulting in:
I ( x) = I (0)e x (1.44)

where it is assumed that in a homogeneous absorber the attenuation coefficient

is uniform and independent of absorber thickness x.
1.6.2. Characteristic absorber thicknesses
Equation (1.44) represents the standard expression for the exponential
attenuation of a monoenergetic narrow photon beam. A typical exponential plot
of intensity I(x) against absorber thickness x of Eq.(1.44) is shown in Fig.1.4 for
a monoenergetic and narrow photon beam. The figure also defines three special
absorber thicknesses used for characterization of photon beams: half-value layer
(HVL), mean free path (MFP) and tenth-value layer (TVL):
HVL (or x1/2) is defined as the thickness of a homogeneous absorber that
attenuates the narrow beam intensity I(0) to one half (50%) of the original
intensity, i.e. I(x1/2)=0.5I(0). The relationship between the HVL x1/2 and
the attenuation coefficient is determined from the basic definition of the
HVL as follows:
I ( x 1/2 ) = 0.5I (0) = I (0)e x 1/2 (1.45)

resulting in:
= e x 1/2


x 1/2 = ln 2 = 0.693



HVL = x 1/2 =

ln 2


MFP (or x ) or relaxation length is the thickness of a homogeneous

absorber that attenuates the beam intensity I(0) to 1/e=0.368 (36.8%) of
its original intensity, i.e. I( x )=0.368I(0). The photon MFP is the average
distance a photon of energy h travels through a given absorber before
undergoing an interaction. The relationship between the MFP x and the
attenuation coefficient is determined from the basic definition of the
MFP as follows:
I ( x ) = I (0) = 0.368 I (0) = I (0)e x (1.47)

resulting in:
= e x


x = 1

MFP = x =


TVL (or x1/10) is the thickness of a homogeneous absorber that attenuates

the beam intensity I(0) to one tenth (10%) of its original intensity, i.e.
I(x1/10)=0.1I(0). The relationship between the TVL x1/10 and the attenuation
coefficient is determined from the basic definition of the TVL as follows:
I ( x 1/10 ) = 0.1I (0) = I (0)e x 1/10 (1.49)

resulting in:
= e x 1/10



x 1/10 = ln 10 = 2.303


TVL = x 1/10 =

ln 10


From Eqs(1.46), (1.48) and (1.50), the linear attenuation coefficient may
be expressed in terms of x1/2, x and x1/10, respectively, as follows:

ln 2 1 ln 10
= =
x 1/2 x x 1/10

resulting in the following relationships among the characteristic thicknesses:

x 1/2 = (ln 2)x =

ln 2
= 0.301x 1/10 (1.52)
ln 10 1/10


FIG.1.4. Intensity I(x) against absorber thickness x for a monoenergetic photon beam.
Half-value layer x1/2, mean free path x and tenth-value layer x1/10 are also illustrated. The
area under the exponential attenuation curve from x=0 to x= is equal to the product I(0) x
where I(0) is the initial intensity of the monoenergetic photon beam. The slope of the tangent
to the attenuation curve at x=0 is equal to I(0) and this tangent crosses the abscissa (x) axis
at x = x .



1.6.3. Attenuation coefficients

In addition to the linear attenuation coefficient , three other related
attenuation coefficients are in use for describing photon beam attenuation
characteristics in absorbers: mass attenuation coefficient m, atomic attenuation
coefficient a and electronic attenuation coefficient e . The attenuation
coefficients are related as follows:
= m = n  a = Z n  e (1.53)


is the mass density of the absorber;

is the number of atoms Na per volume V of the absorber,
i.e. n  = N a / V , and N a / V = N a / m = N A / A with m the mass of the
absorber, NA Avogadros number of atoms per mole and A the atomic mass
of the absorber in grams per mole;
Z is the atomic number of the absorber;

and Z n  is the number of electrons per unit volume of absorber,

i.e. Z n  = ZN A / A.
In radiation dosimetry, two energy-related coefficients are in use: (i) the
energy transfer coefficient tr that accounts for the mean energy transferred E tr
from photons to charged particles (electrons and positrons) in a photonatom
interaction; and (ii) the energy absorption coefficient ab that accounts for the
mean energy absorbed E ab in the medium. The two coefficients are given as
tr =

E tr


E ab



ab =



The light charged particles (electrons and positrons) released or produced

in the absorbing medium through various photon interactions will either:
(a) Deposit their energy to the medium through Coulomb interactions with
orbital electrons of the absorbing medium (collision loss also referred to as
ionization loss); or
(b) Radiate their kinetic energy away in the form of photons through Coulomb
interactions with the nuclei of the absorbing medium (radiation loss).
Typical examples of the mass attenuation coefficient / are shown in
Fig.1.5 with plots of / against photon energy h (in solid dark curves) for
carbon and lead in the energy range from 0.001 to 1000 MeV. Carbon with Z=6
is an example of a low Z absorber; lead with Z=82 is an example of a high
Z absorber. Comparing the two absorbers, it can be noted that at intermediate
photon energies (around 1 MeV), carbon and lead have a similar / of about
0.1 cm2/g. On the other hand, at low photon energies, the / of lead significantly
exceeds the / of carbon, and at energies above 10 MeV, the / of carbon is
essentially flat while the / of lead increases with increasing energy.
1.6.4. Photon interactions on the microscopic scale
The general trends in / depicted in Fig.1.5 reflect the elaborate
dependence of / on the energy h of the photon and the atomic number Z of
the absorber. In penetrating an absorbing medium, photons may experience
various interactions with the atoms of the medium. On a microscopic scale, these
interactions involve either the nuclei of the absorbing medium or the orbital
electrons of the absorbing medium:
(a) Photon interactions with the nucleus of the absorber atoms may be direct
photonnucleus interactions (photonuclear reaction) or interactions
between the photon and the electrostatic field of the nucleus (nuclear pair
(b) Photon interactions with orbital electrons of absorber atoms are
characterized as interactions between the photon and either a loosely bound
electron (Compton effect, triplet production) or a tightly bound electron
(photoelectric effect, Rayleigh scattering).
A loosely bound electron is an electron whose binding energy EB is much
smaller in comparison with the photon energy h, i.e. EB h. An interaction
between a photon and a loosely bound electron is considered to be an interaction
between a photon and a free (i.e. unbound) electron.




FIG.1.5. Mass attenuation coefficient / against photon energy h in the range from 1 keV
FIG. 1.5. Mass attenuation coefficient / against photon energy h in the range from 1 keV
to 1000 MeV for carbon (a) and lead (b). In addition to the total coefficients /, the individual
to 1000 MeV for carbon (a) and lead (b). In addition to the total coefficients /, the
the photoelectric
and pair
for the effect,
and pair production (including triplet production) are also shown. Data Institute
are from the
Science and
of Science
and Technology (NIST).

As far as the photon fate after the interaction with an atom is concerned, there are two
possible outcomes: (i) the photon disappears and is absorbed completely (photoelectric effect,
A tightly
bound triplet
is an electron
EB is
pair production,
and (ii)
the photon
larger its
or slightly
than (Rayleigh
the photon
energy or
For part
a of
scattered andto,changes
but keeps
its energy
photon interaction to occur with a tightly bound electron, the binding energy
The most
with smaller
atoms of
the the
EB of the electron
be of thephoton
order interactions
of, but slightly
photonare: the
Compton effect, photoelectric effect, nuclear pair production, electronic pair production
i.e. EB h. An
interaction between a photon and a tightly bound electron
(triplet production)
and photonuclear reactions. In some of these interactions, energetic
and the atomeffect,
as a whole.
are released frombetween
Compton effect, triplet
production) and electronic vacancies are left in absorber atoms; in other interactions, a portion
of the incident photon energy is used to produce free electrons and positrons. All of these


As far as the photon fate after the interaction with an atom is concerned,
there are two possible outcomes: (i) the photon disappears and is absorbed
completely (photoelectric effect, nuclear pair production, triplet production,
photonuclear reaction) and (ii) the photon is scattered and changes its direction
but keeps its energy (Rayleigh scattering) or loses part of its energy (Compton
The most important photon interactions with atoms of the absorber are:
the Compton effect, photoelectric effect, nuclear pair production, electronic pair
production (triplet production) and photonuclear reactions. In some of these
interactions, energetic electrons are released from absorber atoms (photoelectric
effect, Compton effect, triplet production) and electronic vacancies are left in
absorber atoms; in other interactions, a portion of the incident photon energy
is used to produce free electrons and positrons. All of these light charged
particles move through the absorber and either deposit their kinetic energy in the
absorber (dose) or transform part of it back into radiation through production of
bremsstrahlung radiation.
The fate of electronic vacancies produced in photon interactions with
absorber atoms is the same as the fate of vacancies produced in electron capture
and internal conversion. As alluded to in Section 1.4.11, an electron from a higher
atomic shell of the absorber atom fills the electronic vacancy in a lower shell and
the transition energy is emitted either in the form of a characteristic X ray (also
called a fluorescence photon) or an Auger electron and this process continues
until the vacancy migrates to the outer shell of the absorber atom. A free electron
from the environment will eventually fill the outer shell vacancy and the absorber
ion will revert to a neutral atom in the ground state.
A vacancy produced in an inner shell of an absorber atom migrates to
the outer shell and the migration is accompanied by emission of a series of
characteristic photons and/or Auger electrons. The phenomenon of emission
of Auger electrons from an excited atom is called the Auger effect. Since each
Auger transition converts an initial single electron vacancy into two vacancies,
a cascade of low energy Auger electrons is emitted from the atom. These low
energy electrons have a very short range in tissue but may produce ionization
densities comparable to those produced in an particle track.
The branching between a characteristic photon and an Auger electron is
governed by the fluorescence yield which, as shown in Fig.1.6, for a given
electronic shell, gives the number of fluorescence photons emitted per vacancy
in the shell. The fluorescence yield can also be defined as the probability of
emission of a fluorescence photon for a given shell vacancy. Consequently, as
also shown in Fig.1.6, (1 ) gives the probability of emission of an Auger
electron for a given shell vacancy.


Auger electrons. The phenomenon of emission of Auger electrons from an excited atom is
called the Auger effect. Since each Auger transition converts an initial single electron vacancy
into two vacancies, a cascade of low energy
Auger 1electrons is emitted from the atom. These
low energy electrons have a very short range in tissue but may produce ionization densities
comparable to those produced in an particle track.

FIG.1.6. Fluorescence yields , and M, against atomic number Z of the absorber. Also
FIG. 1.6. Fluorescence yields KK, LL and M
, against atomic number Z of the absorber. Also
shown are
are probabilities
probabilities for
for the
the Auger
Auger effect,
effect, given
given as
as (1
(1 ).
). Data
Data are
Institute of Science and Technology (NIST).
The branching between a characteristic photon and an Auger electron is governed by
yield which,
1.6.5. Photoelectric
effect as shown in Fig. 1.6, for a given electronic shell, gives the
number of fluorescence photons emitted per vacancy in the shell. The fluorescence yield
can also be defined as the probability of emission of a fluorescence photon for a given shell
In Consequently,
the photoelectric
the photoeffect),
as alsoeffect
in Fig. 1.6, (1
) gives
the probability of
interacts with a tightly bound orbital electron of an absorber atom, the photon

disappears and the orbital electron is ejected from the atom as a so-called
photoelectron, with a kinetic energy EK given as:

1.6.5. Photoelectric effect

In the photoelectric effect (sometimes called the photoeffect), the photon

bound orbital electron of an absorber atom, the photon disappears and
E K = hwith

h is the incident photon energy;
and EB is the binding energy of the ejected photoelectron.
A general diagram of the photoelectric effect is provided (seeFig.1.9(a)).
For the photoelectric effect to happen, the photon energy h must exceed
the binding energy EB of the orbital electron to be ejected and, moreover, the
closer h is to EB, the higher the probability of the photoelectric effect happening.
The photoelectric mass attenuation coefficient / is plotted in Fig.1.5 for carbon
and lead as one of the grey curves representing the components of the total /
attenuation coefficient. The sharp discontinuities in the energy h are called


absorption edges and occur when h becomes equal to the binding energy EB of a
given atomic shell. For example, the K absorption edge occurs at h=88 keV in
lead, since the K shell binding energy EB in lead is 88 keV. Absorption edges for
carbon occur at h < 1 keV and, thus, do not appear in Fig.1.5(a).
As far as the photoelectric attenuation coefficient dependence on photon
energy h and absorber atomic number Z is concerned, the photoelectric atomic
attenuation coefficient a goes approximately as Z5/(h)3, while the photoelectric
mass attenuation coefficient / goes approximately as Z4/(h)3.
As evident from Fig.1.5, the photoelectric attenuation coefficient /
is the major contributor to the total attenuation coefficient / at relatively low
photon energies where h is of the order of the K shell binding energy and less
than 0.1 MeV. At higher photon energies, first the Compton effect and then pair
production become the major contributors to the photon attenuation in the absorber.
1.6.6. Rayleigh (coherent) scattering
In Rayleigh scattering (also called coherent scattering), the photon
interacts with the full complement of tightly bound atomic orbital electrons of
an absorber atom. The event is considered elastic in the sense that the photon
loses essentially none of its energy h but is scattered through a relatively
small scattering angle . A general diagram of Rayleigh scattering is given
Since no energy transfer occurs from photons to charged particles, Rayleigh
scattering plays no role in the energy transfer attenuation coefficient and energy
absorption coefficient; however, it contributes to the total attenuation coefficient
/ through the elastic scattering process. The Rayleigh atomic attenuation
coefficient aR is proportional to Z2/(h)2 and the Rayleigh mass attenuation
coefficient R/ is proportional to Z/(h)2.
As a result of no energy transfer from photons to charged particles in the
absorber, Rayleigh scattering is of no importance in radiation dosimetry. As far
as photon attenuation is concerned, however, the relative importance of Rayleigh
scattering in comparison to other photon interactions in tissue and tissue
equivalent materials amounts to only a few per cent of the total / but it should
not be neglected.
1.6.7. Compton effect (incoherent scattering)
The Compton effect (also called incoherent scattering or Compton
scattering) is described as an interaction between a photon and a free as well
as stationary electron. Of course, the interacting electron is not free, rather it is
bound to a nucleus of an absorbing atom, but the photon energy h is much larger


than the binding energy EB of the electron (EB h), so that the electron is said
to be loosely bound or essentially free and stationary.
In the Compton effect, the photon loses part of its energy to the recoil
(Compton) electron and is scattered as a photon h' through a scattering angle ,
as shown schematically in Fig.1.7. In the diagram, the interacting electron is at
the origin of the Cartesian coordinate system and the incident photon is oriented
in the positive direction along the abscissa (x) axis. The scattering angle is the
angle between the direction of the scattered photon h' and the positive abscissa
axis while the recoil angle is the angle between the direction of the recoil
electron and the positive abscissa axis. A general diagram of the Compton effect
is given (seeFig.1.9(c)).

FIG.1.7. Schematic diagram of the Compton effect in which an incident photon of energy
h=1 MeV interacts with a free and stationary electron. A photon with energy h=0.505 MeV
is produced and scattered with a scattering angle =60.

Considerations of conservation of energy and momentum result in the

following three equations for the Compton effect:
(a) Conservation of energy:
h + mec 2 = h '+ mec 2 + E K



h = h '+ E K



(b) Conservation of momentum on the abscissa (x) axis:

pv =

h '
cos +


cos (1.58)


(c) Conservation of momentum on the ordinate (y) axis:


h '
sin +


sin (1.59)


mec2 is the rest energy of the electron (0.511 MeV);
EK is the kinetic energy of the recoil (Compton) electron;
is the velocity of the recoil (Compton) electron;
and c is the speed of light in a vacuum (3 108 m/s).
From equations describing conservation of energy (Eq.(1.57)) and
conservation of momentum (Eqs(1.58) and (1.59)), the basic Compton equation
(also referred to as the Compton wavelength-shift equation) can be derived and is
expressed as follows:
' = =

(1 cos ) = C (1 cos ) (1.60)
m ec

is the wavelength of the incident photon (c/);
is the wavelength of the scattered photon (c/ );
is the wavelength shift in Compton effect ( );
and c, defined as c = h /(mec) = 2 c /(mec 2 ) = 0.024 , is the so-called Compton
wavelength of the electron.
From the Compton equation (Eq.(1.60)), it is easy to show that the
scattered photon energy h and the recoil electron kinetic energy EK depend



on the incident photon energy h as well as on the scattering angle and are,
respectively, given as:
h '(h , ) = h

1 + (1 cos )

E KC (h , ) = h h ' = h h

(1 cos )
= h
1 + (1 cos )
1 + (1 cos )

where is the incident photon energy h normalized to electron rest energy mec2,
i.e. = h /(mec 2 ).
Using Eq.(1.61), it is easy to show that energies of forward-scattered
photons (=0), side-scattered photons ( = /2) and backscattered photons
(=) are in general given as follows:
h ' =0 = h (1.63)
h ' = =




h ' = =

1 + 2

For very large incident photon energies (h ), they are given as:
h ' =0 = h
h ' = = mec 2



h ' = =


m ec 2


From the conservation of momentum equations (Eqs(1.58) and (1.59)),

the following expression for the relationship between the scattering angle and
recoil electron angle can be derived:
cot = (1 + )tan


tan =




Since the range of is from 0 (forward-scattering) through /2 (sidescattering) to (backscattering), it is noted that the corresponding range of is
from =/2 at =0 through to =(1 + )1 for =/2 to =0 at =.
The Compton electronic attenuation coefficient e C steadily decreases
with increasing h from a theoretical value of 0.665 1024 cm2/electron (known
as the Thomson cross-section) at low photon energies to 0.21 1024 cm2/electron
at h =1 MeV, 0.51 1024 cm2/electron at h=10 MeV, and
0.008 1024 cm2/electron at h=100 MeV.
Since Compton interaction is a photon interaction with a free electron, the
Compton atomic attenuation coefficient a C depends linearly on the absorber
atomic number Z, while the electronic coefficient e C and the mass coefficient
C/ are essentially independent of Z. This independence of Z can be observed in
Fig.1.5, showing that C/ for carbon (Z=6) and lead (Z=82) at intermediate
photon energies (~1 MeV), where Compton effect predominates, are equal to
about 0.1 cm2/electron irrespective of Z.
Equation (1.62) gives the energy transferred from the incident photon
to the recoil electron in the Compton effect as a function of the scattering
angle . The maximum energy transfer to recoil electron occurs when the photon
is backscattered (=) and the Compton maximum energy transfer fraction
(fC)max is then given as:
( f C ) max =

(E KC ) max
1 + 2

The mean energy transfer in the Compton effect f C is determined by

normalizing (to incident photon energy h) the mean energy transferred to the
Compton electron E KC . This quantity is very important in radiation dosimetry
and is plotted against incident photon energy h in the Compton graph presented
in Fig.1.8. The figure shows that the fractional energy transfer to recoil electrons

h in the Compton graph presented in Fig. 1.8. The figure shows that the fractional energy
transfer to recoil electrons is quite low at low photon energies ( f C = 0.02 at h = 0.01 MeV)
and then slowly rises through f C = 0.44 atCHAPTER
h = 1 MeV
to reach f C = 0.80 at h = 100 MeV
and approaches one asymptotically at very high incident photon energies.

FIG.1.8. Maximum and mean fractions of incident photon energy h transferred to the recoil

FIG. 1.8. Maximum

h transferred
to the recoil
electron in theand
Data are
the National
of Science
and Technology
electron in
the Compton effect. Data are from the National Institute of Science and
Technology (NIST).
1.6.8. Pair production

is quite low at low photon energies ( f C = 0.02 at h=0.01 MeV) and then slowly
f C =photon
0.44 atenergy
h=1 hMeV
to reach
at h=100
2mfeCc2= =0.80
MeV, with
c2 being the
one asymptotically
at verythe
incident photon
rest energy
of the electron
and positron,
of anenergies.
electronpositron pair in

conjunction with a complete absorption of the incident photon by the absorber atom becomes
Pair production
For the effect to occur, three quantities must be conserved: energy,
charge and momentum. To conserve the linear momentum simultaneously with total energy
h itexceeds
with electric
and charge, theWhen
in free
can only
in theMeV,
the rest (atomic
energy of
the electron
and positron,
field of a mcollision
or orbital
can take ofupana suitable
ec being partner
pair in conjunction
a complete
of the incident
fraction ofelectronpositron
the momentum carried
by the photon.
types of
pair production
are known:
photon by the absorber atom becomes energetically possible. For the effect to
occur, three quantities must be conserved: energy, charge and momentum. To
conserve the linear momentum simultaneously with total energy and charge,
the effect cannot occur in free space; it can only occur in the Coulomb electric
field of a collision partner (atomic nucleus or orbital electron) that can take up
a suitable fraction of the momentum carried by the photon. Two types of pair
production are known:
If the collision partner is an atomic nucleus of the absorber, the pair
production event is called nuclear pair production and is characterized by
a photon energy threshold slightly larger than two electron rest masses
(2mec2=1.022 MeV).



but is
in the

If the
an atomic nucleus
the pair
is called
pair production
is characterized
by aThe
field of
an orbital
electron and
of an
absorber atom.
is called
larger than
electron or
(2mec2 = 1.022
pair two
and MeV).
its threshold photon
2nonetheless possible, is pair production in the Coulomb field of an
energy is 4mec =2.044 MeV.
orbital electron of an absorber atom. The event is called electronic pair production or
triplet production and its threshold photon energy is 4mec2 = 2.044 MeV.

FIG.1.9. Schematic diagrams of the most important modes of photon interaction with atoms of

FIG. 1.9. Schematic diagrams of the most important modes of photon interaction with atoms
an absorber: (a) photoelectric effect; (b) Rayleigh scattering; (c) Compton effect; (d) nuclear
of an
absorber: (a) photoelectric effect; (b) Rayleigh scattering; (c) Compton effect; (d)
pair production; and (e) electronic pair production (triplet production).
pair production; and (e) electronic pair production (triplet production).
The two pair production attenuation coefficients, despite having different origins,
45 are
usually dealt with together as one parameter referred to as pair production. The component
that the nuclear pair production contributes usually exceeds 90%. Nuclear pair production and


The two pair production attenuation coefficients, despite having different

origins, are usually dealt with together as one parameter referred to as pair
production. The component that the nuclear pair production contributes usually
exceeds 90%. Nuclear pair production and electronic pair/triplet production are
shown schematically in Figs 1.9(d) and (e), respectively.
The probability of pair production is zero for photon energy below the
value and
photon energy
the threshold.
in above
Figs 1.9(d)
and 1.9(e),
The pair production atomic attenuation coefficient a and the pair production
probability coefficient
of pair production
is zero
for photon energy
threshold value
/ vary
as Z2 below
and Z,therespectively,
and where
Z is the atomic number of the absorber.

attenuation coefficient a k and the pair production mass attenuation coefficient / vary
Z, respectively,of
Z is the effects
atomic number of the absorber.
as Z2 andpredominance
1.6.9. Relative
1.6.9. Relative predominance of individual effects

As is evident from the discussion above, photons have several options for
with from
atoms. Five
of the
most have
As is evident
the discussion
for interaction
Five of theinmost
electronic are
in Fig.
1.9. Nuclear
are usually
combined and treated under the
the header
pair production.
header pair
probability for a photon to undergo any one of the various interaction
The probability for a photon to undergo any one of the various interaction phenomena
phenomena with an absorber depends on the energy h of the photon and the
with an absorber depends on the energy h of the photon and the atomic number Z of the
atomic number Z of the absorber. In general, the photoelectric effect predominates
absorber. In general, the photoelectric effect predominates at low photon energies, the
at intermediate
and pair
at low
energies, the
at intermediate
and pair
at high photon
production at high photon energies.

Representation of the relative predominance of the three main processes of

1.10. Representation
of the relative predominance of the three main processes of photon
an absorber
the photoelectriceffect,
effect, Compton
Compton effect
interaction with an with
atom: atom:
the photoelectric

46 Figure 1.10 shows the regions of relative predominance of the three most important
individual effects with h and Z as parameters. The two curves display the points on the (h,
Z) diagram for which C = at low photon energies and for which C = for high photon
energies and, thus, delineate regions of photoelectric effect predominance at low photon
energies, Compton effect predominance at intermediate photon energies and pair production


Figure 1.10 shows the regions of relative predominance of the three most
important individual effects with h and Z as parameters. The two curves display
the points on the (h, Z) diagram for which C= at low photon energies and for
which C= for high photon energies and, thus, delineate regions of photoelectric
effect predominance at low photon energies, Compton effect predominance at
intermediate photon energies and pair production predominance at high photon
energies. Figure 1.10 also indicates how the regions of predominance are affected
by the absorber atomic number. For example, a 100 keV photon will interact with
a lead absorber (Z=82) predominantly through the photoelectric effect and with
soft tissue (Zeff 7.5) predominantly through the Compton effect. A 10 MeV
photon, on the other hand, will interact with lead predominantly through pair
production and with tissue predominantly through the Compton effect.
1.6.10. Macroscopic attenuation coefficients
For a given photon energy h and absorber atomic number Z, the
macroscopic attenuation coefficient and energy transfer coefficient tr are
given as a sum of coefficients for individual photon interactions discussed above
(photoelectric, Rayleigh, Compton and pair production):

( + a R + a C + a)
A a


tr =

+ ( a C ) tr + a tr = A a f PE + a C f C + a f PP (1.73)

A a tr

where all parameters are defined in sections dealing with the individual
microscopic effects.
It should be noted that in Rayleigh scattering there is no energy transfer to
charged particles.
The energy absorption coefficient ab (often designated en in the literature)
is derived from tr of Eq.(1.73) as follows:
ab = en = tr (1 g ) (1.74)

where g is the mean radiation fraction accounting for the fraction of the
mean energy transferred from photons to charged particles and subsequently
lost by charged particles through radiation losses. These losses consist of two


components: the predominant bremsstrahlung loss and the small, yet not always
negligible, in-flight annihilation loss.
1.6.11. Effects following photon interactions with absorber
and summary of photon interactions
In the photoelectric effect, Compton effect and triplet production, vacancies
are produced in atomic shells of absorber atoms through the ejection of orbital
electrons from atomic shells. For the diagnostic range and megavoltage range
of photons used for diagnosis and treatment of disease with radiation, the shell
vacancies occur mainly in inner atomic shells and are followed by characteristic
radiation or Auger electrons, the probability of the former given by fluorescence
yield (seeFig.1.6).
Pair production and triplet production are followed by the annihilation
of the positron with a free electron producing two annihilation quanta, most
commonly with an energy of 0.511 MeV each and emitted at 180 from each
other to satisfy conservation of energy, momentum and charge.
ATTIX, F.H., Introduction to Radiological Physics and Radiation Dosimetry, Wiley, New York
CHERRY, S.R., SORENSON, J.A., PHELPS, M.E., Physics in Nuclear Medicine, 3rd edn,
Saunders, Philadelphia, PA (2003).
EVANS, R.D., The Atomic Nucleus, Krieger Publishing, Malabar, FL (1955).
HENDEE, W., RITENOUR, E.R., Medical Imaging Physics, 4th edn, Wiley, New York (2002).
JOHNS, H.E., CUNNINGHAM, J.R., The Physics of Radiology, 3rd edn, Thomas, Springfield,
IL (1984).
KHAN, F., The Physics of Radiation Therapy, 4th edn, Lippincott, Williams and Wilkins,
Baltimore, MD (2009).
KRANE, K., Modern Physics, 3rd edn, Wiley, New York (2012).
PODGORSAK, E.B., Radiation Physics for Medical Physicists, 2nd edn, Springer, Heidelberg,
New York (2010).
ROHLF, J.W., Modern Physics from to Z0, Wiley, New York (1994).


Department of Surgery and Cancer,
Faculty of Medicine,
Imperial College London,
London, United Kingdom
Division of Human Health,
International Atomic Energy Agency,
Radiobiology is the study (both qualitative and quantitative) of the actions of
ionizing radiations on living matter. Since radiation has the ability to cause changes
in cells which may later cause them to become malignant, or bring about other
detrimental functional changes in irradiated tissues and organs, consideration of
the associated radiobiology is important in all diagnostic applications of radiation.
Additionally, since radiation can lead directly to cell death, consideration of the
radiobiological aspects of cell killing is essential in all types of radiation therapy.
At the microscopic level, incident rays or particles may interact with
orbital electrons within the cellular atoms and molecules to cause excitation or
ionization. Excitation involves raising a bound electron to a higher energy state,
but without the electron having sufficient energy to leave the host atom. With
ionization, the electron receives sufficient energy to be ejected from its orbit and
to leave the host atom. Ionizing radiations (of which there are several types) are,
thus, defined through their ability to induce this electron ejection process, and

Present address: Department of Radiology, Leiden University Medical Centre, Leiden,




the irradiation of cellular material with such radiation gives rise to the production
of a flux of energetic secondary particles (electrons). These secondary particles,
energetic and unbound, are capable of migrating away from the site of their
production and, through a series of interactions with other atoms and molecules,
give up their energy to the surrounding medium as they do so.
This energy absorption process gives rise to radicals and other chemical
species and it is the ensuing chemical interactions involving these which are the
true causatives of radiation damage. Although the chemical changes may appear to
operate over a short timescale (~105s), this period is nonetheless a factor of ~1018
longer than the time taken for the original particle to traverse the cell nucleus. Thus,
on the microscopic scale, there is a relatively long period during which chemical
damage is inflicted (Table2.1).
It is important to note that, irrespective of the nature of the primary
radiation (which may be composed of particles and/or electromagnetic waves),
the mechanism by which energy is transferred from the primary radiation beam to
biological targets is always via the secondary electrons which are produced. The
initial ionization events (which occur near-instantaneously at the microscopic level)
are the precursors to a chain of subsequent events which may eventually lead to the
clinical (macroscopic) manifestation of radiation damage.
Expression of cell death in individual lethally damaged cells occurs
later, usually at the point at which the cell next attempts to enter mitosis. Gross
(macroscopic and clinically observable) radiation effects are a result of the
wholesale functional impairment that follows from lethal damage being inflicted to
large numbers of cells or critical substructures. The timescale of the whole process
may extend to months or years. Thus, in clinical studies, any deleterious health
effects associated with a radiation procedure may not be seen until long after the
diagnostic test or treatment has been completed (Table2.1).

Approximate timescale

Initial ionizing event

1018 s

Transit of secondary electrons

1015 s

Production of ion radicals

1010 s

Production of free radicals

109 s

Chemical changes

105 s

Individual cell death

Gross biological effects





2.3.1. Types of ionizing radiation
In nuclear medicine, there are four types of radiation which play a relevant
role in tumour and normal tissue effects: gamma () radiation, beta () radiation,
alpha ()particles and Auger electrons. Gamma radiation
Gamma radiation is an electromagnetic radiation of high energy
(usually above 25keV) and is produced by subatomic particle interactions.
Electromagnetic radiation is often considered to be made up of a stream of
wave-like particle bundles (photons) which move at the speed of light and whose
interaction properties are governed mainly by their associated wavelength.
Although the collective ionization behaviour of large numbers of photons can
be predicted with great accuracy, individual photon interactions occur at random
and, in passing through any type of matter, a photon may interact one or more
times, or never. In each interaction (which will normally involve a photoelectric
event, a Compton event or a pair production event), secondary particles are
produced, usually electrons (which are directly ionizing) or another photon
of reduced energy which itself can undergo further interactions. The electrons
undergo many ionizing events relatively close to the site of their creation and,
therefore, contribute mostly to the locally absorbed dose. Any secondary photons
which may be created carry energy further away from the initial interaction site
and, following subsequent electron-producing interactions, are responsible for
the dose deposition occurring at sites which are more distant from the original
interaction. Beta radiation
Beta radiation is electrons emitted as a consequence of radionuclide decay.
A decay process can occur whenever there is a relative excess of neutrons ()
or protons (+). One of the excess neutrons is converted into a proton, with the
subsequent excess energy being released and shared between an emitted electron
and an anti-neutrino. Many radionuclides exhibit decay and, in all cases, the
emitted particle follows a spectrum of possible energies rather than being emitted
with a fixed, discrete energy. In general, the average energy is around one third
of the maximum energy. Most emitting radionuclides also emit photons as a
consequence of the initial decay, leaving the daughter nucleus in an excited,
metastable state. Since particles are electrons, once ejected from the host atom,


they behave exactly as do the electrons created following the passage of a ray,
giving up their energy (usually of the order of several hundred kiloelectronvolts)
to other atoms and molecules through a series of collisions.
For radionuclides which emit both particles and photons, it is usually
the particulate radiation which delivers the greatest fraction of the radiation dose
to the organ which has taken up the activity. For example, about 90% of the dose
delivered to the thyroid gland by 131I arises from the component. On the other
hand, the emissions contribute more significantly to the overall whole body
dose. Alpha particles
Alpha radiation is emitted when heavy, unstable nuclides undergo decay.
Alpha particles consist of a helium nucleus (two protons combined with two
neutrons) emitted in the process of nuclear decay. The particles possess
approximately 7000 times the mass of a particle and twice the electronic charge,
and give up their energy over a very short range (<100m). Alpha particles
usually possess energies in the megaelectronvolt range, and because they lose this
energy in such a short range are biologically very efficacious, i.e. they possess a
high linear energy transfer (LET; see Section 2.6.3) and are associated with high
relative biological effectiveness (RBE; see Section 2.6.4). Auger electrons
Radionuclides which decay by electron capture or internal conversion
leave the atom in a highly excited state with a vacancy in one of the inner shell
electron orbitals. This vacancy is rapidly filled by either a fluorescent transition
(characteristic X ray) or non-radiative (Auger) transition, in which the energy
gained by the electron transition to the deeper orbital is used to eject another
electron from the same atom. Auger electrons are very short range, low energy
particles that are often emitted in cascades, a consequence of the inner shell atomic
vacancy that traverses up through the atom to the outermost orbital, ejecting
additional electrons at each step. This cluster of very low energy electrons can
produce ionization densities comparable to those produced by an particle track.
Thus, radionuclides which decay by electron capture and/or internal conversion
can exhibit high LET-like behaviour close (within 2 nm) to the site of the decay.




Radiation induced damage to biological targets may result from direct or
indirect action of radiation (Fig.2.1):
Direct action involves ionization or excitation (via Coulomb interactions) of
the atoms in the biological target. This gives rise to a chain of events which
eventually leads to the observable (macroscopic) damage. In normally
oxygenated mammalian cells, the direct effect accounts for about one third
of the damage for low LET radiations such as electrons and photons.
Indirect action involves radiation effects on atoms or molecules which
are not constituent parts of the biological target. Since cells exist in a rich
aqueous environment, the majority of indirect actions involve the ionization
or excitation of water molecules. The free radicals subsequently created
may then migrate and damage the adjacent biological targets. Indirect action
is the main cause of radiation damage and, in normoxic cells, accounts for
about two thirds of the damage.
Indirect action is predominant with low LET radiation, e.g. X and rays,
while direct action is predominant with high LET radiation, e.g. particles and

FIG.2.1. Illustration of the difference between direct and indirect damage to cellular DNA.



2.4.1. Role of oxygen

Radiation effects may be influenced by several factors, especially the
presence or absence of oxygen. The free radicals (denoted by a dot placed to the
right of the atomic symbol) produced as a result of direct or indirect effects are
very reactive and seek to interact with other molecules which can share or donate
electrons. Molecular oxygen (O2) has two unpaired electrons and readily reacts
with free radicals, causing an increased likelihood that DNA (deoxyribonucleic
acid) will be damaged by the indirect process. Important reactions via which
oxygen can increase biological damage are:
R i + O 2 RO 2 i (highly toxic )
H i + O 2 HO 2 i
HO 2 i + HO 2 i = H 2O 2 (highly toxic) + O 2

where R represents an organic molecule.

The oxygen enhancement ratio (OER) is given by the dose in hypoxia (total
absence of oxygen) divided by the dose in air required to achieve an equivalent
biological effect. For low LET radiation, such as rays, the OER has a value
of ~3. For high LET radiation, such as particles, the OER decreases to almost
2.4.2. Bystander effects
Bystander effects occur when a cell which has not been traversed by a
charged particle is damaged as a result of radiation interactions occurring in
neighbouring cells. The discovery of the bystander effect poses a challenge to
the traditional view that all radiation damage stems from direct interactions of
charged particles with critical cellular targets. For this reason, it still remains
controversial in radiobiology. A possible explanation is that irradiated cells may
send out a stress signal to nearby cells, which may elicit a response, e.g. the
initiation of apoptosis, in those cells. The overall relevance of the bystander
effect is presently difficult to gauge. It is probably most significant in radiation
protection considerations involving low doses since it amplifies the overall
radiation effect in situations where not all of the cells in a tissue are subjected
to particle transversal, i.e. the overall radiation risk to that tissue is higher than
would be expected from consideration of the gross response exhibited by those
cells which have been directly traversed by charged particles.




2.5.1. DNA damage
DNA damage is the primary cause of cell death caused by radiation.
Radiation exposure produces a wide range of lesions in DNA such as single
strand breaks (SSBs), double strand breaks (DSBs), base damage, proteinDNA
cross-links and proteinprotein cross-links (seeFig.2.1). The number of DNA
lesions generated by irradiation is large, but there are a number of mechanisms
for DNA repair. As a result, the percentage of lesions causing cell death is very
small. The numbers of lesions induced in the DNA of a cell by a dose of 12 Gy
are approximately: base damages: >1000; SSBs: ~1000; DSBs: ~40. DSBs play
a critical role in cell killing, carcinogenesis and hereditary effects. There are
experimental data showing that the initially produced DSBs correlate with
radiosensitivity and survival at lower dose, and that unrepaired or misrepaired
DSBs also correlate with survival after higher doses. Furthermore, there is
experimental evidence for a causal link between the generation of DSBs and the
induction of chromosomal translocations with carcinogenic potential.
2.5.2. DNA repair
DNA repair mechanisms are important for the recovery of cells from
radiation and other damaging agents. There are multiple enzymatic mechanisms
for detecting and repairing radiation induced DNA damage. DNA repair
mechanisms, such as base excision repair, mismatch repair and nucleotide
excision repair, respond to damage such as base oxidation, alkylation and strand
intercalation. Excision repair consists of cleavage of the damaged DNA strand
by enzymes that cleave the polynucleotide chain on either side of the damage,
and enzymes which cleave the end of a polynucleotide chain allowing removal
of a short segment containing the damaged region. DNA polymerase can then
fill in the resulting gap using the opposite undamaged strand as a template. For
DSBs, there are two primary repair pathways, non-homologous end joining
(NHEJ) and homologous recombination. NHEJ repair operates on blunt ended
DNA fragments. This process involves the repair proteins recognizing lesion
termini, cleaning up the broken ends of the DNA molecule, and the final ligation
of the broken ends. DSB repair by homologous recombination utilizes sequence
homology with an undamaged copy of the broken region and, hence, can only
operate in late S/G2-phases of the cell cycle. Undamaged DNA from both strands
is used as a template to repair the damage. In contrast to NHEJ, the repair process
of homologous recombination is error-free. Repair by NHEJ operates throughout
the cell cycle but dominates in G1/S-phases. The process is error-prone because it


does not rely on sequence homology. Unrepaired or misrepaired damage to DNA

will lead to mutations and/or chromosome damage in the exposed cell. Mutations
might lead to cancer or hereditary effects (when germ cells are exposed), whereas
severe chromosome damage often leads to cell death.
2.6.1. Concept of cell death
Radiation doses of the order of several grays may lead to cell loss. Cells
are generally regarded as having been killed by radiation if they have lost
reproductive integrity, even if they have physically survived. Loss of reproductive
integrity can occur by apoptosis, necrosis, mitotic catastrophe or by induced
senescence. Although all but the last of these mechanisms ultimately results in
physical loss of the cell, this may take a significant time to occur.
Apoptosis or programmed cell death can occur naturally or result from
insult to the cell environment. Apoptosis occurs in particular cell types after low
doses of irradiation, e.g. lymphocytes, serous salivary gland cells, and certain
cells in the stem cell zone in testis and intestinal crypts.
Necrosis is a form of cell death associated with loss of cellular membrane
activity. Cellular necrosis generally occurs after high radiation doses.
Reproductive cell death is a result of mitotic catastrophe (cells attempt
to divide without proper repair of DNA damage) which can occur in the first
few cell divisions after irradiation, and it occurs with increasing frequency after
increasing doses.
Ionizing radiation may also lead to senescence. Senescent cells are
metabolically active but have lost the ability to divide.
2.6.2. Cell survival curves
A quantitative understanding of many aspects of biological responses to
radiation may be made by consideration of the behaviour of the underlying cell
survival (dose response) characteristics. Although the practical determination of
cell survival curves is potentially fraught with experimental and interpretational
difficulties and is best performed by persons who are experts in such procedures,
an appreciation of the structure and meaning of such curves, even in a purely
schematic context, can be very helpful in understanding the role played by the
various factors which influence radiation response.
Figure2.2 shows the typical shape of a cell survival curve for mammalian
tissue. Physical radiation dose is plotted on the linear horizontal axis while


fractional cell survival is plotted on the logarithmic vertical axis. Each of the
individual points on the graph represents the fractional survival of cells resulting
from delivery of single acute doses of the specified radiation, which in this case
is assumed to be radiation. (In the context of the subject, an acute dose of
radiation may be taken to mean one which is delivered at high dose rate, i.e. the
radiation delivery is near instantaneous.) Mammalian cell survival curves plotted
in this way are associated with two main characteristics: a finite initial slope (at
zero dose) and a gradually increasing slope as dose is increased.

Surviving fraction

Dose (Gy)

FIG.2.2. A radiation cell survival curve plots the fraction of plated cells retaining colony
forming ability (cell surviving fraction) versus radiation absorbed dose.

2.6.3. Dose deposition characteristics: linear energy transfer

As noted above, the energy transfer to the absorbing medium (whether
that be animate or inanimate material) is via secondary electrons created by the
passage of the primary ionizing particle or ray. LET is a measure of the linear
rate at which radiation is absorbed in the absorbing medium by the secondary
particles and is defined by the International Commission on Radiation Units
and Measurements (ICRU) as being the quotient dE/dl, where dE is the average
energy locally imparted to the medium by a charged particle of specified energy
in traversing a distance dl. The unit usually employed for LET is kiloelectronvolt
per micrometre and some representative values are listed in Table2.2.



Radiation type

Linear energy transfer (keV/m)



250kVp X rays


10MeV protons


2.5MeV particles


1MeV electrons


10keV electrons


1keV electrons


Co rays

For radiobiological studies in particular, the concept of LET is problematic

since it relates to an average linear rate of energy deposition but, at the
microscopic level (i.e. at dimensions comparable with the critical cellular targets),
the energy deposited per unit length along different parts of a single track may
vary dramatically. In particular, as charged particles lose energy in their passage
through a medium via the result of collision and ionizing processes, the LET rises
steeply to its highest value towards the very end of their range. The change in
LET value along the track length is one reason why average LET values correlate
poorly with observed (i.e. macroscopic) biological effects. For these reasons, the
directly measured RBE is of much greater use as an indicator of the differing
biological efficacies of various radiation types.
2.6.4. Determination of relative biological effectiveness
For a given biological end point, the RBE of the high LET radiation is
defined as the ratio of the isoeffective doses for the reference (low LET) and the
high LET radiation (Fig.2.3). The reference radiation is usually 60Co g rays or
high energy (250kVp) X rays.


Surviving fraction


Dose (Gy)
FIG.2.3. The relative biological effectiveness of a radiation is defined as the ratio of the
dose required to produce the same reduction in cell survival as a reference low linear energy
transfer (LET) radiation.

If the respective low and high LET isoeffective doses are dL and dH, then:


If the basic cell survival curves are described in terms of the

linearquadratic (LQ) model, then the surviving fraction S as a function of acute
doses at low and high LET is respectively given as:
S L = exp( L d L L d L2 ) (2.2)
S H = exp( H d H H d H
) (2.3)

where the suffixes L and H again respectively refer to the low and high LET
Figure2.4 shows an example of how the RBEs determined at any particular
end point (cell surviving fraction) vary with changing dose for a given radiation
fraction size for a low LET radiation. The maximum RBE (RBEmax) occurs
at zero dose and, in terms of microdosimetric theory, corresponds to the ratio


between the respective high and low LET linear radiosensitivity constants, H
and L, i.e.:




RBE max =

FIG.2.4. Relative biological effectiveness (RBE) as a function of the radiation dose per

If the quadratic radiosensitivity coefficients (H and L) are unchanged with

changing LET (i.e. H=L), then, at high doses, the RBE tends to unity. However,
this constancy of , assumed by the theory of Kellerer and Rossi, has been
challenged and, if does change with LET, then RBE will tend asymptotically to
an alternativeminimum value (RBEmin) given by:
RBE min =


and the working RBE at any given dose per fraction is given as:


+ 4d L RBE min
( / )L RBE max + ( / )L RBE max
( / )L + d L (2.6)
2 ( / )L + d L



when expressed in terms of the low LET dose per fraction dL or:


( / )L + ( / )L + 4d H ( / )L RBE max + RBE min
d H

2d H


when expressed in terms of the high LET dose per fraction dH.
Figure2.4 was derived using RBEmax =5, RBEmin=1 and (/)L=3 Gy,
but the general trend of a steadily falling RBE with increasing dose per fraction
is independent of the chosen values. Clearly, the assumption of a fixed value
of RBE, if applied to all fraction sizes, could lead to gross clinical errors and
Eqs(2.6) and (2.7) make the point that determination of RBEs in a clinical setting
is potentially complex and will depend on accurate knowledge of RBEmax and (if
it is not unity) RBEmin. Although there is not yet clear evidence over whether
or not there is a consistent trend for RBEmin to be non-unity, the possibility is
nevertheless important as it may hold very significant implications.
Figure2.4 also shows schematically how the rate of change of RBE with
changing dose per fraction is influenced by the existence of a non-unity RBEmin
parameter. Even for a fixed value of RBEmax, the potential uncertainty in the RBE
values at the fraction sizes likely to be used clinically might themselves be very
large if RBEmin is erroneously assumed to be unity. These uncertainties would be
compounded if there were an additional linkage between RBEmax and the tissue
/ value.
As is seen from Eqs(2.6) and (2.7), the RBE value at any particular dose
fraction size will also be governed by the low LET / ratio (a tissue dependent
parameter which provides a measure of how tissues respond to changes in dose
fractionation) and the dose fraction size (a purely physical parameter) at the point
under consideration. Finally, and as has been shown through the earlier clinical
experience with neutron therapy, the RBEmax value may itself be tissue dependent,
likely being higher for the dose-limiting normal tissues than for tumours. This
tendency is borne out by experimental evidence using a variety of ion species as
well as by theoretical microdosimetric studies. This potentially deleterious effect
may be offset by the fact that, in carbon-, helium- and argon-ion beams, LET
(and, hence, RBE) will vary along the track in such a way that it is low at the
entry point (adjacent to normal tissues) and highest at the Bragg peak located
in the tumour. However, although this might be beneficial, it does mean that the
local RBE is more spatially variable than is indicated by Eq.(2.6).
Owing to the difficulties in setting reference doses at which clinical
inter-comparisons could be made more straightforward, Wambersie proposed
that a distinction be made between the reference RBE and the clinical


RBE. Thus, the reference RBE might be that determined at 2 Gy fractions on

a biological system end point representative, for example, of the overall late
tolerance of normal tissues. As more clinical experience of using the particular
radiation becomes available, a more practical clinical RBE evolves, this being
the reference RBE empirically weighted by collective clinical experience and
by volume effects related to the beam characteristics, geometry or technical
2.6.5. The dose rate effect and the concept of repeat treatments
When mammalian cells are irradiated, it is helpful to visualize their
subsequent death as resulting from either of two possible processes. In the first
process, the critical nuclear target (DNA) is subjected to a large deposition of
energy which physically breaks both strands of the double helix structure and
disrupts the code sufficiently to disallow any opportunity of repair. This process
can be thought of as a single-hit process and the total amount of DNA damage
created this way is directly proportional to the dose delivered.
In the second process, an ionizing event occurs and releases only sufficient
energy to disrupt the coding carried by one strand of the DNA. Following this
event, and if the irradiation continues, two outcomes are possible: either the
broken strand will restore itself to its original state (no lethality) or, prior to full
repair taking place, a second, independent radiation event may occur in the same
location and damage the opposite strand of the DNA, a complementary action
between the two damaged strands then leading to cell lethality in what is called a
two-hit process. Since this route depends on there being two independent events,
each having a probability proportional to dose, the number of damaged DNA
targets created this way is proportional to dose dose, i.e. dose2. Once created,
the radiation damage due to these two possible routes is indistinguishable
(i.e. both processes are lethal). From this simplified description, it is clear that the
observed radiation response characterized in the cell survival curve will consist
of two components: one linear with dose and the other quadratic, i.e. proportional
to dose2. This phenomenological description qualitatively explains the shape of
a radiation survival curve, with a finite initial slope at low dose followed by an
increasingly downward curvature as dose increases.
However, the amount of damage created in the second process is dependent
on the ability of the second break to be induced before the first break has repaired
itself and, thus, is dependent on the dose rate.
Figure2.5 shows a range of response curves in which the doses are
delivered at four different dose rates, the individual doses taking proportionately
longer to deliver as dose rate is reduced. This graph illustrates that reducing the
dose rate causes the overall shape of the response curve to become less curvy


than in the acute case, but that the initial slope remains unchanged. When the
doses are all delivered at a very low dose rate, as is the case for most radionuclide
therapies, the response is essentially a straight line, when the curves are plotted
on a log-linear scale, as is common practice for radiation survival curves. This
means that the low dose response is purely exponential.

Surviving fraction

Dose (Gy)

FIG.2.5. Surviving fraction as a function of dose for different dose rates. It is important
to note that most radionuclide therapies are delivered at low dose rate in the range of
0.10.5 Gy/h, when the survival curve is almost linear.

2.6.6. The basic linearquadratic model

The basic equation describing the shape of the cell survival curves shown
in Fig.2.2 is referred to as the LQ model, which has a biophysical origin. Cell
survival following delivery of an acute dose d is given as:
S = exp( d d 2 ) (2.8)

where (in units of Gy1) and (in units of Gy2) are the respective linear and
quadratic sensitivity coefficients.



If the treatment is repeated in N well spaced fractions, then the net survival
is SN, where:
S N = S N = exp(N d N d 2 ) (2.9)

Taking natural logarithms on both sides of Eq.(2.9) and dividing throughout

by leads to:
ln S N

= Nd

Nd 2
( / )

2.6.7. Modification to the linearquadratic model

for radionuclide therapies
Targeted radionuclide therapy normally involves irradiation of the
tumour/normal tissues at a dose rate which is not constant but which reduces as
treatment proceeds, as a consequence of the combination of radionuclide decay
and biological clearance of the radiopharmaceutical. To allow for this, a more
extensive formulation of the LQ model is required.
2.6.8. Quantitative intercomparison of different treatment types
In many aspects of LQ modelling, a term called the biological effective
dose (BED) is employed to assess and inter-compare different treatment types.
BED is defined as:

ln S N

= Nd 1 +
( / )

Although the parameters and are rarely known in detail for individual
tumours or tissues, values of the ratio / (in units of grays) are becoming
increasingly known from clinical and experimental data. In general, / is
systematically higher (520 Gy) for tumours than for critical, late-responding
normal tissues (25 Gy) and it is this difference which provides the BED concept
with much of its practical usefulness.



For non-acute treatments (those in which the dose delivery is protracted

over a long time period on account of a lower dose rate), the BED is re-written

d g (t )
BED = Nd 1 +
( / )

where g(t) is a function of the time t taken for delivery:

g (t ) =

2 1 exp( t )

and where is the mono-exponential time constant relating to the repair of

sublethal damage. is related to the tissue repair half-time (T1/2) via:


For a treatment delivery at constant dose rate R, the delivered dose d is

related to treatment time t via d=Rt, thus:

2 R 1 exp( t )
BED = Rt 1 +

( / )

When t > 12 h, Eq.(12.15) simplifies to:

2 R
BED = Rt 1 +
( / )

2.6.9. Cellular recovery processes

At lower doses and dose rates, cellular recovery may play an important role
in the fixation of the radiation damage. There are three broad types of cellular
radiation damage:
(a) Lethal damage in which the cellular DNA is irreversibly damaged to such
an extent that the cell dies or loses its proliferative capacity;
(b) Sublethal damage in which partially damaged DNA is left with sufficient
capacity to restore itself over a period of a few hours, provided there is no
further damage during the repair period;


(c) Potentially lethal damage in which repair of what would normally be

a lethal event is made possible by manipulation of the post-irradiation
cellular environment.
2.6.10. Consequence of radionuclide heterogeneity
The effectiveness per unit dose of a radiopharmaceutical depends on the
heterogeneity of the radionuclide distribution. Global non-uniformity of a source
distribution, which results in pockets of cells (tumour or normal tissue) receiving
less than the average dose will almost always result in a greater fraction of cell
survivors, than if all cells receive a uniform dose. The one possible exception
would be if a radiopharmaceutical would selectively localize at sensitive target
cells, within an organ, that are key for organ regeneration or function, e.g. crypt
cells in the colon. The cellular response also depends on the microdosimetry,
especially if the radiopharmaceutical in question selectively localizes on the cell
surface or internalizes within a certain cohort of cells within a tumour/normal
organ. Radiolabels that selectively localize on the surface of cells or are
internalized may exhibit geometric enhancement factors that modulate a response.
The reader is referred to ICRU Report 67 on absorbed dose specification in
nuclear medicine for more details.
2.7.1. Classification of radiation damage (early versus late)
Cells which are lethally affected by radiation may continue to function
for some time after the infliction of the damage, only dying when attempting
to undergo subsequent cell division (mitosis). Clinically observed radiation
effects in whole tissues or organs reflect the damage inflicted to large numbers
of constituent cells and, thus, appear on a timescale which is governed largely
by the underlying proliferation rates of those cells. Such observable effects
are classified as being either late or early, depending on the speed at which
they manifest themselves following irradiation. Late effects appear months or
years after irradiation and appear in structures which proliferate very slowly,
e.g. kidney. Early (or acute) effects appear within days, weeks or months of
irradiation and are associated with fast-proliferating epithelial tissues, e.g. bone
marrow, mucosa, intestinal tract, etc.



In most types of radiotherapy, it is the late effects which are considered

to be most critical and which generally limit the total dose which may be
delivered to the tumour. If the radiation tolerance of the late-responding tissues is
exceeded, then the subsequent reactions, depending on the tissues in which they
arise, may seriously affect mobility and/or quality of life, and may even be life
threatening. Such problems arise long after the completion of treatment and are,
thus, impossible to correct. These are the serious considerations which are at the
heart of the therapeutic index concept discussed below (seeSection 2.7.3). Acute
reactions in radiotherapy, although they may be unpleasant, are usually transient
and easier to control by adjustment of the treatment dose delivery pattern and/or
simple medication. In radionuclide therapies, it is in most instances possible
to circumvent acute radiation toxicities once they begin to occur, such as by
accelerating clearance of the radiopharmaceutical. Chronic toxicities, such as
to the kidney, usually occur at times which are long relative to the lifetime of
the radionuclide. Hence, considerable importance should be attributed to the
administration of safe activities of therapeutic radionuclides that do not exceed
any dose limiting constraints.
2.7.2. Determinants of tumour response
Irrespective of the mechanism used to achieve tumour targeting, the
potential advantage of radionuclide therapy over other forms of radiation therapy
is its ability to deliver dose to both the local disease and to occult tumour deposits.
In nuclear medicine, the primary determinants of treatment effectiveness
The tumour specificity of the radionuclide carrier.
The homogeneity of uptake of the carrier within the targeted tumour(s).
The intrinsic RBE (seeSection 2.6.4) of the radiation used for the therapy:
this is determined primarily by the nature of the radionuclide emissions
(e.g. particles, particles, low energy rays, Auger electrons, etc.).
The range of the particles, as determined by their energies.
The total dose delivered.
The responsiveness of the targeted tumour cells to radiation. This will be
determined by radiobiological properties such as cellular radiosensitivity
and the variations of sensitivity within the cell cycle, the oxygen status of
the cells (fully oxic, partially oxic or hypoxic), the ability of the cells to
recover from sublethal radiation damage and the degree to which tumour
growth (repopulation) may occur during therapy.



These factors are complementary and interactive, and should not

be considered in isolation from each other. Thus, for example, significant
non-uniformity of uptake within the tumour may result in dose cold spots, but
the detrimental potential of these might be offset by the selection of a radionuclide
which emits particles of sufficient range to produce a cross-fire effect within the
cold spots from those adjacent cells which are properly targeted. The significance
of cold spot and cross-fire effects is further dependent on the size of the tumour
deposit under consideration.
2.7.3. The concept of therapeutic index in radiation
therapy and radionuclide therapy
The therapeutic index of a particular radiation treatment (often referred to in
older publications as the therapeutic ratio) is a measure of the resultant damage
to the tumour vis a vis the damage to critical normal structures. Treatments with
a high therapeutic index will demonstrate good tumour control and low normal
tissue morbidity; treatments with a low therapeutic index will be associated with
a low tumour control and/or high morbidity. There have been several attempts to
provide quantitative definitions of therapeutic index, but it is usually sufficient to
consider therapeutic index as being a qualitative concept any new treatment
which, relative to an existing treatment, improves tumour control and/or reduces
morbidity is said to be associated with an improved therapeutic index.
In conventional (external beam) radiotherapy, the normal tissues at risk
will be those immediately adjacent to the tumour being treated. Doses to the
normal tissues (along with the risk of toxicity) may be reduced by attention to
a combination of physical and radiobiological factors. In targeted radionuclide
therapy, the tumour may be single and discrete (as is the case in most external
beam therapy) or may consist of distributed masses or metastatic deposits at
several locations within the body. The normal tissues at risk may themselves be
widely distributed but, more particularly, may be a reflection of the particular
uptake pattern of the targeting compound being used for the therapy.
2.7.4. Long term concerns: stochastic and deterministic effects
The radiation detriment which results from radiation exposure may be
classified as being either stochastic or deterministic in nature. Stochastic effects
(e.g. hereditary damage, cancer induction) are those for which the likelihood
of them occurring is dose related, but the severity of the resultant condition is
not related to the dose received. Deterministic effects (e.g. cataract induction,
general radiation syndromes, bone marrow ablation, etc.) manifest themselves
with a severity which is dose related. In general, it is predominantly stochastic


effects which need to be considered as potential side effects from diagnostic uses
of radionuclides, although deterministic damage may result if the embryo or fetus
is irradiated. For radionuclide therapy applications, the concerns relate to both
stochastic and deterministic effects.
2.8.1. Radionuclide targeting
Tumour targeted radiotherapy is a very promising approach for the
treatment of wide-spread metastasis and disseminated tumour cells. This
technique aims to deliver therapeutic irradiation doses to the tumour while
sparing normal tissues by targeting a structure that is abundant in tumour cells,
but rare in normal tissues. This can be done by using antibodies labelled with
a therapeutic relevant radionuclide acting against a specific tumour target.
Radiolabelled antibody therapy has already become common in the treatment of
non-Hodgkins lymphoma, e.g. 131I-tositumomab (Bexxar) and 90Y-ibritumomab
tiuxetan (Zevalin), and exhibits great potential for being extended to other
diseases. A good example is epidermal growth factor (EGF) labelled with 125I
which will bind EGF receptors. EGF receptors are overexpressed on tumour cells
in many malignancies such as highly malignant gliomas. At present, several other
radiolabelled antibodies are being used in experimental models and in clinical
trials to study their feasibility in other types of cancer.
2.8.2. Whole body irradiation
Conventional external beam radiotherapy involves controlled irradiation of
a carefully delineated target volume. Normal structures adjacent to the tumour
will likely receive a dose, in some cases a moderately high dose, but the volumes
involved are relatively small. The rest of the body receives only aminimal dose,
mostly arising from radiation scattered within the patient from the target volume
and from a small amount of leakage radiation emanating from the treatment
machine outside the body.
Targeted radionuclide therapies are most commonly administered
intravenously and, thus, can give rise to substantial whole body doses and, in
particular, doses to the radiation sensitive bone marrow. Once the untargeted
activity is removed from the blood, it may give rise to substantial doses in
normal structures, especially the kidneys. Furthermore, the activity taken up by



the kidneys and targeted tumour deposits may (if ray emissions are involved)
continue to irradiate the rest of the body.
2.8.3. Critical normal tissues for radiation and radionuclide therapies
Since the radiation doses used in radionuclide therapies are much higher
than the doses used for diagnosis, (prolonged) retention of the pharmaceuticals
within the blood circulation and, hence, increased accumulation of radionuclides
in non-tumour cells, might lead to unwanted toxicities. The bone marrow, kidney
and liver are regarded as the main critical organs for systemic radionuclide therapy.
Other organs at risk are the intestinal tract and the lungs. The bone marrow is
very sensitive towards ionizing radiation. Exposure of the bone marrow with high
doses of radiation will lead to a rapid depression of white blood cells followed
a few weeks later by platelet depression, and in a later stage (approximately one
month after exposure) also by depression of the red blood cells. In general, these
patients could suffer from infections, bleeding and anaemia. Radiation damage to
the gastrointestinal tract is characterized by de-population of the intestinal mucosa
(usually between 3 and 10 days) leading to prolonged diarrhoea, dehydration,
loss of weight, etc. The kidneys, liver and lungs will show radiation induced
damage several months after exposure. In kidneys, a reduction of proximal tubule
cells is observed. These pathological changes finally lead to nephropathy. In the
liver, hepatocytes are the radiosensitive targets. Since the lifespan of the cells is
about a year, deterioration of liver function will become apparent between 3 and
9 months after exposure. In lungs, pulmonary damage is observed in two waves:
an acute wave of pneumonitis and later fibrosis.
The determinants of normal tissue response from radionuclide studies
is a large subject due to the diversity of radiopharmaceuticals with differing
pharmacokinetics and biodistribution, and the widely differing responses and
tolerances of the critical normal tissues. A principal determinant of the type of
toxicity depends on the radionuclide employed. For example, isotopes of iodine
localize in the thyroid (unless blocked), salivary glands, stomach and bladder.
Strontium, yttrium, samarium, fluorine, radium, etc. concentrate in bone.
Several radiometals, such as bismuth, can accumulate in the kidney. If these
radionuclides are tightly conjugated to a targeting molecule, the biodistribution
and clearance are determined by that molecule. For high molecular weight
targeting agents, such as an antibody injected intravenously, the slow plasma
clearance results in marrow toxicity being the principal dose limiting organ. For
smaller radiolabelled peptides, renal toxicity becomes of concern. When studying
a new radiopharmaceutical or molecular imaging agent, it is always important
to perform a detailed study of the biodistribution at trace doses, to ensure the



absence of radionuclide sequestration within potentially sensitive tissue, such as

the retina of the eye or the germ cells of the testes.
A review of normal tissue toxicities resulting from radionuclide therapies is
given by Meredith et al. (2008).
2.8.4. Imaging the radiobiology of tumours
The development of molecular imaging using positron emission tomography
(PET) has given rise to new radiotracers which have the potential to assess
several features of radiobiological relevance for therapy planning. One tracer
that is becoming more widely available for PET imaging is fluorothymidine.
This radiotracer exhibits the property of becoming selectively entrapped within
cells that are progressing through S-phase (DNA replication) of the cell cycle,
thus providing a signal which should be proportional to cell proliferation,
andminimizing the signal from cells in G0 or in cell cycle arrest. The ability to
selectively identify only replicating cells separate from all tumour cells present
within the computed tomography-determined tumour volume may present an
excellent opportunity for more accurate measures of the initial viable tumour
burden as well as evaluating tumour response. Complementary to measuring
tumour response is the measurement of therapeutic efficacy through radiotracers
that selectively target cell death. Radiotracers are under development with the
ability to selectively bind to receptors expressed on cells undergoing programmed
cell death, e.g. radiolabelled annexin V. Another area of active research is in the
field of hypoxia imaging. Cells within a tumour microenvironmental region of low
partial oxygen pressure, i.e. hypoxia, are known to exhibit a great radio-resistance
to both radiation and chemotherapy relative to those under normoxic conditions.
A number of PET radiotracers are under evaluation for imaging tumour
hypoxia with PET, including fluoromisonidazole (18F-FMISO), fluoroazomycin
arabinoside (18F-FAZA) and copper-diacetyl-bis(N4-methylthiosemicarbazone)
(64Cu-ATSM). The ability to measure the radiobiological attributes of a tumour
prior to therapy may provide invaluable information concerning the relative
resistance/aggressiveness of tumours, leading to improved management of these
2.8.5. Choice of radionuclide to maximize therapeutic index
The choice of the optimum radionuclide to maximize the therapeutic index
depends on a number of factors. First, the range of the emitted particles from the
radionuclide should depend on the type of tumour being treated. For leukaemia
or micrometastatic deposits, consisting of individual or small clusters of tumour
cells, there is a distinct advantage of using radionuclides which emit very short


range particles. Since particles have ranges of <100 m in tissue, particle

emitters would have an advantage, if the targeting molecule were able to reach
all tumour cells. However, particle emitting radionuclides are not widely
available and are extremely expensive. In addition, the short range of particles
can be a disadvantage for bulk tumours. For these reasons, almost all therapeutic
radionuclides utilized in the clinic today consist of medium (131I) or long range
(90Y, 186Re) emitters. These radionuclides are advantageous when treating solid
tumours for which target receptor (antigen) expression may be heterogeneous, or
with non-uniform delivery, due to the greater cross-fire range of their emissions
(ranging up to a 1cm range in unit density tissue).
A second important consideration is the choice of radionuclide half-life.
If the half-life is too short, then the radiolabelled tumour targeting agent may
have insufficient time to reach its target, resulting in aminimal therapeutic index.
Increasing the half-life will increase the therapeutic index, but render the patient
radioactive for a longer period of time, resulting in prolonged confinement, greater
expense and radiation risks to staff and family. Pure emitting radionuclides such
as 90Y and 32P have advantages in that theyminimize the exposure to personnel
assisting the patient. The half-life of the radionuclide should ideally match the
biological uptake and retention kinetics of the tumour-targeting carrier used. For
large protein carriers such as antibodies, radionuclides with half-lives of several
days are required to optimize the therapeutic index. For smaller molecular
targeting agents such as peptides, short lived radionuclides may be better suited
tominimize radioactive waste.
Thirdly, it is necessary to consider radiochemistry, ease and stability of the
radiolabelled end product. All of these factors need to be taken into consideration
in order to produce the optimum therapeutic targeting compound for use in
clinical therapeutic applications.
DALE, R.G., JONES, B. (Eds), Radiobiological Modelling in Radiation Oncology, The British
Institute of Radiology, London (2007).
HALL, E.J., GIACCIA, A.J., Radiobiology for the Radiologist, 6th edn, Lippincott, Williams
and Wilkins, Philadelphia, PA (2006).
in Nuclear Medicine, Rep. 67, Nuclear Technology Publishing, Ashford, United Kingdom
MEREDITH, R., WESSELS, B., KNOX, S., Risks to normal tissue from radionuclide therapy,
Semin. Nucl. Med. 38 (2008) 347357.


Department of Diagnostic Radiology,
Uddevalla Hospital,
Uddevalla, Sweden
Division of Radiation, Transport and Waste Safety,
International Atomic Energy Agency,
Medical exposure is the largest human-made source of radiation exposure,
accounting for more than 95% of radiation exposure. Furthermore, the use of
radiation in medicine continues to increase worldwide more machines are
accessible to more people, the continual development of new technologies and
new techniques adds to the range of procedures available in the practice of
medicine, and the role of imaging is becoming increasingly important in day
to day clinical practice. The introduction of hybrid imaging technologies, such
as positron emission tomography/computed tomography (PET/CT) and single
photon emission computed tomography (SPECT)/CT, means that the boundaries
between traditional nuclear medicine procedures and X ray technologies are
becoming blurred. Worldwide, the total number of nuclear medicine examinations
is estimated to be about 35 million per year.
In Chapter2, basic radiation biology and radiation effects were described,
demonstrating the need for a system of radiation protection. Such a system allows
the many beneficial uses of radiation to be utilized, but at the same time ensures
detrimental radiation effects are either prevented orminimized. This can be
achieved by having the objectives of preventing the occurrence of deterministic
effects and of limiting the probability of the stochastic effects to a level that is
considered acceptable. In a nuclear medicine facility, consideration needs to
be given to the patient, the staff involved in performing the nuclear medicine
procedures, members of the public and other staff that may be in the nuclear
medicine facility, carers and comforters of patients undergoing procedures,



and persons who may be undergoing a nuclear medicine procedure as part of a

biomedical research project.
This chapter discusses how the objectives stated above are achieved
through a system of radiation protection, and how such a system should be
applied practically in a hospital in general and in nuclear medicine specifically.
The means for achieving the objectives of radiation protection have evolved
over many years to the point where, for some time, there has been a reasonably
consistent approach throughout the world namely the system of radiological
protection, as espoused by the International Commission on Radiological
Protection (ICRP). The following will briefly describe this system, specifically as
it applies to nuclear medicine.
3.2.1. The International Commission on Radiological
Protection system of radiological protection
The principles of radiation protection and safety upon which the IAEA
safety standards are based are those developed by the ICRP. The detailed
formulation of these principles can be found in ICRP publications and they
cannot easily be paraphrased without losing their essence. However, a brief,
although simplified, summary of the principles is given in this section.
The ICRP recommends a system of radiological protection to cover all
possible exposure situations. There are many terms associated with the system
and some of these will now be introduced.
The ICRP in its Publication 103 [3.1] divides all possible situations
of where exposure can occur into three types planned exposure situations,
emergency exposure situations and existing exposure situations. For the practice
of nuclear medicine, only the first situation is relevant. The use of radiation
in nuclear medicine is a planned exposure situation it needs to be under
regulatory control, with an appropriate authorization in place from the regulatory
body before operation can commence. Misadministration, spills and other such
incidents or accidents can give rise to what is called potential exposure, but these
remain part of the planned exposure situation as their occurrence is considered
in the granting of an authorization. It should be noted that the ICRP has used the
term practice to describe a planned exposure situation such as the operation of a
nuclear medicine facility.



The ICRP then puts exposure of individuals into three categories medical
exposure, occupational exposure and public exposure:
Medical exposure refers primarily to exposure incurred by patients for
the purpose of medical diagnosis or treatment. It also refers to exposures
incurred by individuals helping in the support and comfort of patients
undergoing diagnosis or treatment, and by volunteers in a programme of
biomedical research involving their exposure.
Occupational exposure is the exposure of workers incurred in the course of
their work.
Public exposure is exposure incurred by members of the public from all
exposure situations, but excluding any occupational or medical exposure.
All three need to be considered in the nuclear medicine facility.
An individual person may be subject to one or more of these categories of
exposure, and for radiation protection purposes such exposures are dealt with
The ICRP system has three fundamental principles of radiological
protection, namely:
The principle of justification: Any decision that alters the radiation exposure
situationshould do more good than harm.
The principle of optimization of protection: The likelihood of incurring
exposures, the number of people exposed and the magnitude of their
individual doses should all be kept as low as reasonably achievable
(ALARA), taking into account economic and societal factors.
The principle of limitation of doses: The total dose to any individual
from regulated sources in planned exposure situations other than medical
exposure of patients should not exceed the appropriate limits recommended
by the ICRP. Recommended dose limits are given in Table3.1.
In a nuclear medicine facility, occupational and public exposures are subject
to all three principles, whereas medical exposure is subject to the first two only.
More detail on the application of the ICRP system for radiological protection as
it applies to a nuclear medicine facility is given in the remainder of this chapter.




Type of limit
Effective dose
Annual equivalent dose in:
Lens of the eyed
Skine, f
Hands and feet



20 mSv per year, averaged over defined

periods of 5 yearsb

1 mSv in a yearc

20 mSv
500 mSv
500 mSv

15 mSv
50 mSv

Limits on effective dose are for the sum of the relevant effective doses from external
exposure in the specified time period and the committed effective dose from intakes of
radionuclides in the same period. For adults, the committed effective dose is computed for a
50 year period after intake, whereas for children it is computed for the period up to reaching
70 years of age.
With the further provision that the effective dose should not exceed 50 mSv in any single
year. Additional restrictions apply to the occupational exposure of pregnant women.
In special circumstances, a higher value of effective dose could be allowed in a single year,
provided that the average over 5 years does not exceed 1 mSv/a.
In 2011, the ICRP recommended that the occupational dose limit be lowered from the
previous 150 mSv/a to 20 mSv/a, averaged over 5 years, and with no more than 50 mSv in
any single year.
The limitation on effective dose provides sufficient protection for the skin against stochastic
Averaged over a 1cm2 area of skin regardless of the area exposed.

3.2.2. Safety standards

Safety standards are based on knowledge of radiation effects and on the
principles of protection described above. In this respect, the development of
safety standards by the IAEA follows a well established approach. The United
Nations Scientific Committee on the Effects of Atomic Radiation (UNSCEAR),
a body set up by the United Nations in 1955, compiles, assesses and disseminates
information on the health effects of radiation and on levels of radiation exposure
due to different sources; this information was taken into account in developing
the standards. Following a decision made in 1960, the IAEA safety standards
are based on the recommendations of the ICRP, which also take account of the
scientific information provided by UNSCEAR.
Purely scientific considerations, however, are only part of the basis for
decisions on protection and safety, and the safety standards implicitly encourage


decision makers to make value judgements about the relative importance of

different kinds of risks and about the balancing of risks and benefits. General
acceptance of risk is a matter of consensus and, therefore, international safety
standards should provide a desirable international consensus for the purpose of
For these reasons, international consensus is integral to the IAEA safety
standards, which are prepared with the wide participation of and approval by its
Member States and relevant international organizations. The current version of
what is commonly called the Basic Safety Standards (BSS) is entitled Radiation
Protection and Safety of Radiation Sources: International Basic Safety Standards
(2014) [3.2]. The BSS are jointly sponsored by the European Commission,
the Food and Agriculture Organization of the United Nations, the IAEA, the
International Labour Organization, the OECD Nuclear Energy Agency, the
Pan American Health Organization (PAHO), the United Nations Environment
Programme and the World Health Organization (WHO).
The BSS comprises five sections: Introduction, General requirements for
protection and safety, Planned exposure situations, Emergency exposure situations
and Existing exposure situations, as well as four schedules. The purpose of the
BSS is to establish basic requirements for protection against exposure to ionizing
radiation and for the safety of radiation sources that may deliver such exposure.
The requirements of the BSS underpin the implementation of radiation protection
in a nuclear medicine facility, supplemented by the relevant IAEA Safety Guides
and Safety Reports.
3.2.3. Radiation protection quantities and units
The basic dosimetry quantity for use in radiation protection is the mean
organ or tissue dose DT given by:
DT = T / mT


mT is the mass of the organ or tissue T;
and T is the total energy imparted by radiation to that tissue or organ.
The International System of Units (SI) unit of mean organ dose is joules per
kilogram (J/kg) which is termed gray (Gy).
Owing to the fact that different types of ionizing radiation will have
different effectiveness in damaging human tissue at the same dose, and the fact


that the probability of stochastic effects will depend on the tissue irradiated, it is
necessary to introduce quantities to account for these factors. Those quantities
are equivalent dose and effective dose. Since they are not directly measurable,
the International Commission on Radiation Units and Measurements (ICRU)
has defined a set of operational quantities for radiation protection purposes (area
monitoring and personal monitoring): the ambient dose equivalent, directional
dose equivalent and personal dose equivalent.
Regarding internal exposure from radionuclides, the equivalent dose
and the effective dose are not only dependent on the physical properties of the
radiation but also on the biological turnover and retention of the radionuclide.
This is taken into account in the committed dose quantities (equivalent and
effective). Equivalent dose
It is a well known fact in radiobiology that densely ionizing radiation such
as particles and neutrons will cause greater harm to a tissue or organ than
rays and electrons at the same mean absorbed dose. This is because the dense
ionization events will result in a higher probability of irreversible damage to the
chromosomes and a lower chance of tissue repair. To account for this, the organ
dose is multiplied with a radiation weighting factor in order to get a quantity that
more closely reflects the biological effect on the irradiated tissue or organ. This
quantity is called the equivalent dose and is defined as:
H T = wr DT,r (3.2)

DT,R is the mean tissue or organ dose delivered by type R radiation;
and wR is the radiation weighting factor.
For X rays, rays and electrons, wR =1; for particles, wR =20. The SI unit of
equivalent dose is joules per kilogram (J/kg), which is termed sievert (Sv). In a
situation of exposure from different types of radiation, the total equivalent dose is
the sum of the equivalent dose from each type of radiation.



The relationship between the probability of stochastic effects and equivalent
dose is found to depend on the organ or tissue irradiated. To account for this,
tissue weighting factors wT are introduced. They should represent the relative
contribution of an organ or tissue T to the total detriment due to the stochastic
effects resulting from a uniform irradiation of the whole body. The total tissue
weighted equivalent dose is called effective dose and is defined as:



where HT is the equivalent dose in organ or tissue T.

The sum is performed over all organs and tissues of the human body
considered to be sensitive to the induction of stochastic effects. Recommended
tissue weighting factors are found in ICRP Publication 103 [3.1]. Despite
depending on the sex and age of the person, for the purposes of radiation
protection, the values for tissue weighting factors are taken as constants and are
applicable to the average population.
The use of effective dose has many advantages in practical radiation
protection. Very different exposure situations (e.g. internal and external exposure
by different types of radiation) can be combined and result in a single value, the
effective dose. Committed dose
When radionuclides are taken into the body, the resulting dose is received
throughout the period of time during which they remain in the body. The total
dose delivered during this period of time is referred to as the committed dose and
is calculated as a specified time integral of the rate of receipt of the dose. The
committed equivalent dose is defined as:
H T ( ) =

t0 +


H T (t ) dt (3.4)


is the time of intake;

and is the integration time.



For workers and adult members of the general public, is taken to be

50 years while for children 70 years is regarded as appropriate.
The committed effective dose is given by:
E( ) =

T H T ( )

(3.5) Operational quantities

For all types of external radiation, the operational quantities for area
monitoring are defined on the basis of a dose equivalent value at a point in the
ICRU sphere. It is a sphere of tissue-equivalent material (30cm in diameter with
a density of 1 g/cm3 and a mass composition of: 76.2% oxygen, 11.1% carbon,
10.1% hydrogen and 2.6% nitrogen). For radiation monitoring, it adequately
approximates the human body in regards to the scattering and attenuation of the
radiation fields under consideration.
The operational quantities for area monitoring defined in the ICRU
sphere should retain their character of a point quantity. This is achieved by
introducing the terms expanded and aligned radiation field in the definition
of these quantities. An expanded radiation field is a hypothetical field in which
the spectral and the angular fluence have the same values at all points of a
sufficiently large volume equal to the values in the actual field at the point of
interest. The expansion of the radiation field ensures that the whole ICRU sphere
is thought to be exposed to a homogeneous radiation field with the same fluence,
energy distribution and direction distribution as at the point of interest of the real
radiation field. If all radiation is aligned in the expanded radiation field so that it
is opposed to a radius vector specified for the ICRU sphere, the aligned and
expanded radiation field is obtained. In this hypothetical field, the ICRU sphere
is homogeneously irradiated from one direction, and the fluence of the field is
the integral of the angular differential fluence at the point of interest in the real
radiation field over all directions. In the expanded and aligned radiation field, the
value of the dose equivalent at any point in the ICRU sphere is independent of
the direction distribution of the radiation in the real radiation field.
For area monitoring, the operational quantity for penetrating radiation is
the ambient dose equivalent, H*(10):
The ambient dose equivalent at a point in a radiation field is the dose
equivalent that would be produced by the corresponding expanded and
aligned field in the ICRU sphere at a depth of 10mm on the radius vector
opposing the direction of the aligned field.



For area monitoring, the operational quantity for low-penetrating radiation

is the directional dose equivalent H(0.07, ):
The directional dose equivalent at a point in a radiation field is the dose
equivalent that would be produced by the corresponding expanded field in
the ICRU sphere at a depth of 0.07mm in a specified direction .
The personal dose equivalent Hp(d)is defined as:
The equivalent dose at a depth d in soft tissue below a specified point on
the body.
The relevant depth is d=10mm for penetrating radiations (photon energies
above 15keV), while depths d=0.07mm and d=3mm are used for weakly
penetrating radiations (photon energies below 15keV) in skin and the lens of the
eye, respectively.
3.3.1. General aspects
Implementation of radiation protection in a nuclear medicine facility must
fit in with, and be complementary to, the systems for implementing medical
practice in the facility. Radiation protection must not be seen as something
imposed from outside and separate to the real business of providing medical
services and patient care. Most countries have their own radiation protection
legislation and regulatory framework, typically requiring any facility or person
wishing to provide or perform nuclear medicine procedures to have an appropriate
authorization from the radiation protection regulatory body. The requirements to
be fulfilled in order to be granted such an authorization will vary from country to
country, but in general, compliance with the requirements of the BSS would be
To achieve a high standard of radiation protection, the most important thing
is to establish a safety based attitude in every individual, such that protection and
accident prevention are regarded as a natural part of daily duties. This objective is
basically achieved by education and training, and encouraging a questioning and
learning attitude, but also by a positive and cooperative attitude from the national
authorities and the employer in supporting radiation protection with sufficient
resources, both in terms of personnel and money. A feeling of responsibility


can only be achieved if the people involved regard the rules and regulations as
necessary, and are a support to and not a hindrance in their daily work. Every
individual should also know their responsibilities through formal assignment of
3.3.2. Responsibilities Licensee and employer
The licensee of a nuclear medicine facility, through the authorization issued
by the regulatory body, has the prime responsibility for applying the relevant
national regulations and meeting the conditions of the licence. The licensee may
appoint other people to carry out actions and tasks related to these responsibilities,
but the licensee retains overall responsibility. In particular, the nuclear medicine
physician, the medical physicist, the nuclear medicine technologist, the
radiopharmacist and the radiation protection officer (RPO) all have key roles
and responsibilities in implementing radiation protection in a nuclear medicine
facility, and these are discussed in more detail below.
The BSS need to be consulted for details on all of the requirements for
radiation protection that are assigned to licensees. Employers are also assigned
many responsibilities, in cooperation with the licensee, for occupational radiation
protection. Key responsibilities for the licensee include ensuring that the
necessary personnel (nuclear medicine physicians, medical physicists, nuclear
medicine technologists, radiopharmacists and an RPO) are appointed, and that
the individuals have the necessary education, training and competence to perform
their respective duties. Clear responsibilities for personnel must be assigned; a
radiation protection programme (RPP) must be established and the necessary
resources provided; a comprehensive quality assurance (QA) programme must
be established; and education and training of personnel supported. Nuclear medicine specialist
The general medical and health care of the patient is, of course, the
responsibility of the individual physician treating the patient. However, when the
patient presents in the nuclear medicine facility, the nuclear medicine specialist
has the particular responsibility for the overall radiation protection of the patient.
This means responsibility for the justification of a given nuclear medicine
procedure for the patient, in conjunction with the referring medical practitioner,
and responsibility for ensuring the optimization of protection in the performance
of the examination or treatment.


RADIATION PROTECTION Nuclear medicine technologist

The technologist has a key position, and their skill and care to a large extent
determine the optimization of the patients exposure. Radiation protection officer
It is highly recommended that the licensee appoint a person to oversee and
implement radiation protection matters in the hospital. This person is called an
RPO or radiation safety officer. The RPO should have a good theoretical and
practical knowledge of the properties and hazards of ionizing radiation, as well
as protection. In addition, the RPO should possess necessary knowledge of all
the appropriate legislation and codes of practice relating to the uses of ionizing
radiation in the relevant medical area, e.g. nuclear medicine. The RPO, unless
also a qualified medical physicist in nuclear medicine, has no responsibilities for
radiation protection in medical exposure. Medical physicist
The medical physicist is a person who by education and training is
competent to practise independently in one or more of the subfields in
medical physics. For instance, a medical physicist in nuclear medicine should
have a comprehensive knowledge of the imaging equipment used, including
performance specifications, physical limitations of the equipment, calibration,
quality control and image quality. The medical physicist should also be qualified
in handling radiation protection matters associated with nuclear medicine, and
has particular responsibilities for radiation protection in medical exposure,
including the requirements pertaining to imaging (for diagnostic procedures),
calibration, dosimetry and QA. Whenever possible, a medical physicist should
serve as the RPO (seeabove). Other important tasks for the medical physicist
are to be responsible for QA and for the local continuing education in radiation
protection of the nuclear medicine staff and other health professionals. Other personnel
Other personnel that may have responsibilities in radiation protection in
nuclear medicine include radiopharmacists and other staff that may have been
trained to perform special tasks, such as contamination tests or some quality
control tests.



3.3.3. Radiation protection programme

The BSS require a licensee (and employer where appropriate) to develop,
implement and document a protection and safety programme commensurate
with the nature and extent of the risks of the practice to ensure compliance with
radiation protection standards. Such a programme is often called an RPP and
each nuclear medicine facility should have one. The RPP for a nuclear medicine
facility is quite complex as it needs to cover all relevant aspects of protection of
the worker, the patient and the general public. The details of such an RPP can be
found in Ref.[3.3].
For an RPP to be effective, the licensee needs to provide for its
implementation, including the resources necessary to comply with the programme
and arrangements to facilitate cooperation between all relevant parties.
3.3.4. Radiation protection committee
An effective way to supervise compliance with the RPP is the formation of
a committee for radiation protection. Since a representative of the management
is usually a member of the radiation protection committee, communication
with the representative may be the most appropriate. The members of the
radiation protection committee should include an administrator representing the
management, the chief nuclear medicine physician, a medical physicist, the RPO,
a nuclear medicine technologist, possibly a nurse for patients undergoing therapy
with radiopharmaceuticals, and a maintenance engineer.
3.3.5. Education and training
According to the BSS, provision must be made to ensure that all personnel
on whom protection and safety depend are appropriately trained and qualified
so that they understand their responsibilities and perform their duties with
appropriate judgement and according to defined procedures. Such personnel
clearly include the nuclear medicine physician (or other medical specialist
wishing to perform nuclear medicine procedures), nuclear medicine technologist,
medical physicist, radiopharmacist and the RPO. However, there are additional
staff that may also need appropriate training, such as nurses working with
radioactive patients and maintenance staff. Details about appropriate levels of
training are given in Ref.[3.3].




It is an important task for the medical physicist to be actively involved in
the planning and design of the nuclear medicine facility. Factors that are to be
considered are:
Safety of sources;
Optimization of protection for staff and the general public;
Preventing uncontrolled spread of contamination;
Maintaining low background where most needed;
Fulfilment of national requirements regarding pharmaceutical work.
3.4.1. Location and general layout
The location of the nuclear medicine facility within the hospital or clinic
is not critical, but a few factors need to be considered. It should be readily
accessible, especially for outpatients, who constitute the majority of the patients.
The facility should also be located away from radiotherapy sources and other
strong sources of ionizing radiation such as a cyclotron, which can interfere with
the measuring equipment. Isolation wards for patients treated with radionuclides
should be located outside of the nuclear medicine facility.
The general layout of the nuclear medicine facility should take into account
a possible separation of the work areas and the patient areas. It is also essential to
reduce uncontrolled spread of contamination. This will be achieved by locating
rooms for preparation of radiopharmaceuticals as far away as possible from
rooms for measurements and patient waiting areas. Another important factor is to
reduce the transport of unsealed sources within the facility. The general layout is
from a low activity area close to the entrance to high activity areas at the opposite
end. More details regarding floor planning and additional topics can be found in
the IAEAs Nuclear Medicine Resources Manual [3.4]. It should be borne inmind
that the design of facilities is an important tool in the optimization of protection
of workers and the general public. This is further discussed in Section 3.6.2.
3.4.2. General building requirements
The design of the facility should take into consideration the type of work
to be performed and the radionuclides (and their activity) intended to be used.
The ICRPs concept of categorization of hazard can be used in order to determine
the special needs concerning ventilation and plumbing, and the materials used
in walls, floors and work-benches. The different rooms in the facility will be
categorized as low, medium or high hazard areas.


Of special concern in a nuclear medicine facility is the risk of

contamination, and if contamination occurs, the ability to contain it and clean
it up. Therefore, the floors and work-benches should generally be finished in an
impermeable material which is washable and resistant to chemical change, with
all joints sealed. The floor cover should be curved to the wall. The walls should
also be easily cleaned. Chairs and beds used in high hazard areas should be easily
decontaminated. However, some attention has to be given to the comfort of the
patients, for instance in the waiting areas.
Rooms in which unsealed sources, especially radioactive aerosols or gases,
may be produced or handled should have an appropriate ventilation system
that includes a fume hood, laminar air flow cabinet or glove box. It should be
noted that this might also be necessary in the examination room depending on
the radiopharmaceutical used in ventilation scintigraphy. Details regarding fume
hoods, etc. are given in Chapter9.
If the regulatory body allows the release of aqueous waste to the sewer,
a dedicated sink needs to be used, and this needs to be easily decontaminated.
Local rules for the discharge shall be available.
A separate bathroom for the exclusive use by injected patients is
recommended. A sign requesting patients to always sit down, flush the toilet
well and wash their hands should be displayed to lower the risk of contamination
of the floor and to ensure adequate dilution of excreted radioactive materials.
The bathroom should include a sink as a normal hygiene measure and should
be finished in materials that are easily decontaminated. Local rules should be
available for cleaning the toilet. The patient toilet facilities should not be used
by hospital staff as it is likely that the floor, toilet seat and taps will frequently be
Drain-pipes from the nuclear medicine facility should go as directly as
possible to the main building sewer. It should be noted that some countries require
that drain-pipes from a nuclear medicine facility and especially from isolation
wards for patients undergoing radionuclide therapy end up in a delay tank.
3.4.3. Source security and storage
The licensee needs to establish a security system to prevent theft, loss,
unauthorized use or damage to sources. It should be included in all steps from
ordering and delivery of the sources to disposal of spent sources. Only authorized
personnel are permitted to order radionuclides. Routines for delivery and
unpacking shipments should be available, as well as routines for safe handling
and storage of sources. Records of all sources should be kept. The user is always
responsible for the security of sources and, in principle, it should be possible
to identify where an individual source is located or how it has been used, even


if it has left the facility in a patient. The regulatory body should promptly be
informed in cases of lost or stolen sources.
When a radioactive source is not in use, it should always be stored. In a
nuclear medicine facility, the sources are generally stored in the room where
preparation of radiopharmaceuticals is undertaken. Storage of sources is further
discussed in Chapter9.
It is necessary to consider the possible consequences of an accidental fire
and to take steps tominimize the risk of this. Careful selection of non-flammable
construction materials when building the storage facility will greatly reduce this
hazard. The storage facility should not be used to hold any highly flammable or
highly reactive materials. Liaison with the local firefighting authority is necessary
and their advice should be sought regarding provision of firefighting equipment
in the vicinity of the radioactive waste store.
3.4.4. Structural shielding
Structural shielding should be considered in a busy facility where large
activities are handled and where many patients are waiting and examined. In a
PET/CT facility, structural shielding is always necessary and the final design will
generally be determined by the PET application because of the high activities
used and because of the high energy of the annihilation radiation. Careful
calculations should be performed to ensure the need and construction of the
barrier. Such calculations should include not only walls but also the floor and
ceiling, and must be made by a qualified medical physicist. Radiation surveys
should always be performed to ensure the correctness of the calculations.
The correct design of protective barriers is of the utmost importance not
only from a protection but also from an economic point of view. If the basic
calculations are wrong, it will become very expensive to correct the mistakes
later when the whole construction is completed. It is, therefore, very important
that a qualified expert, such as a medical physicist, be consulted in the planning
3.4.5. Classification of workplaces
With regard to occupational exposure, the BSS require the classification of
workplaces as controlled areas or as supervised areas.
In a controlled area, individuals follow specific protective measures to
control radiation exposures. It will be necessary to designate an area as controlled
if it is difficult to predict doses to individual workers or if individual doses
may be subject to wide variations. The controlled area must be delineated and



it is convenient to use existing structural boundaries, which should already be

considered at the planning stage of a facility.
A supervised area is any area for which occupational exposure conditions
are predictable and stable. They are kept under review even though specific
additional protective measures and safety provisions are not normally needed.
In a nuclear medicine facility, the rooms for preparation, storage (including
radioactive waste) and injection of the radiopharmaceuticals will be controlled
areas. Owing to the potential risk of contamination, the imaging rooms and
waiting areas for injected patients might also be classified as controlled areas.
The area housing a patient to whom therapeutic amounts of activity have been
given will also be a controlled area. In the case of pure emitters, such as 90Y,
Sr or 32P, which are not excreted from the body, the area may not need to be
classified as a controlled area.
3.4.6. Workplace monitoring
Workplace monitoring means checking the facility for the presence of
radiation or radioactive contamination. The two basic types of workplace
monitoring are exposure monitoring and contamination monitoring. Exposure
monitoring (sometimes called area monitoring or radiation surveying) consists
of measuring radiation levels (in microsieverts per hour) at various points using
an exposure meter or survey meter. Contamination monitoring is the search for
extraneous radioactive material deposited on surfaces.
Routine workplace monitoring should be performed at predefined places
in the facility as defined by the RPO. It is an advantage if one member of staff is
appointed to take the measurements. The staff member should be well trained in
handling the instrument. The results should be recorded and investigated if they
exceed the investigation levels predefined by the RPO.
More details regarding workplace monitoring are given in Chapters 9
and 20.
3.4.7. Radioactive waste
The radioactive waste in a nuclear medicine facility comprises many
different types of waste. It may be of high activity, such as a technetium generator,
or of low activity, such as from biomedical procedures or research. In addition,
it may have a long or short half-life and it may be in a solid, liquid or gaseous
form. Radioactive waste needs to be safely managed because it is potentially
hazardous to human health and the environment. Through good practices in the
use of radionuclides, the amount of waste can be significantly reduced but not
eliminated. It is important that safe waste management, in full compliance with


all relevant regulations, is considered and planned for at the early stages of any
projects involving radioactive materials. It is the responsibility of the licensee to
provide safe management of the radioactive waste. It should be supervised by the
RPO and local rules should be available.
Containers to allow segregation of different types of radioactive waste
should be available in areas where the waste is generated. The containers must
be suitable for the purpose (volume, shielding, being leakproof, etc.). Each type
of waste should be kept in separate containers that are properly labelled to supply
information about the radionuclide, physical form, activity and external dose rate.
A room for interim storage of radioactive waste should be available. The
room should be locked, properly marked and, if necessary, ventilated. Flammable
waste should be placed separately. It is essential that all waste be properly packed
in order to avoid leakage during storage. Biological waste should be refrigerated
or put in a freezer. Records should be kept, so that the origin of the waste can be
The final disposal of the radioactive waste produced in the nuclear medicine
facility includes several options: storage for decay and disposal as cleared waste
into the sewage system (aqueous waste), through incineration or transfer to a
landfill site (solid waste), or transfer of sources to the vendor or to a special waste
disposal facility outside of the hospital.
For many of the wastes generated in hospitals, storage for decay is a useful
option because the radionuclides generally have short half-lives. This can be
done in the hospital and may include some treatment of the wastes to ensure
safe storage. Other types of waste containing radionuclides with longer half-lives
must be transferred to a special waste treatment, storage and disposal facility
outside of the hospital. One option is to return the source to the vendor. This is an
attractive option for radionuclide generators and might also be useful for sealed
sources used in a quality control programme. The option of returning the source
should be provided for in the purchase process.
For diagnostic patients there is generally no need for collection of excreta.
Ordinary toilets can be used. For therapy patients, there are different policies in
different countries, either to use separate toilets equipped with delay tanks or an
active treatment system, or to allow the excreta to be released directly into the
sewage system. This is further discussed in Chapter20.
Detailed requirements for protection against occupational exposure are
given in Section 3 of the BSS, and recommendations on how to meet these
requirements are given in IAEA Safety Guides [3.53.7]. All of these safety


standards are applicable to nuclear medicine practice, and in addition Ref.[3.3]

provides further specific advice. A summary of the most relevant issues for
nuclear medicine is given in this section.
3.5.1. Sources of exposure
Exposure of workers may arise from unsealed sources either through
external irradiation of the body or through entry of radioactive substances into the
body. The main precautions required in dealing with external irradiation depend
on the physical characteristics of the emitted radiation and the activity as reflected
by the specific dose rate constant as well as the half-life of the radionuclide.
When a radionuclide enters the body, the internal exposure will depend on factors
such as the physical and chemical properties of the radionuclide, the activity and
the biokinetics.
Every type of work performed in a nuclear medicine facility will
make a contribution to the external exposure of the worker: unpacking
radioactive material, activity measurements, storage of sources, preparation of
radiopharmaceuticals, administration of radiopharmaceuticals, patient handling
and examination, care of the radioactive patient and handling of radioactive
waste. Generally, the yearly effective dose to staff working full time in nuclear
medicine with optimized protection should be well below 5 mSv.
Among the different tasks involved, the highest effective dose is received
from the patient at injection and imaging. The dose rate close to the patient can
be quite high, for instance, 300Sv/h at 0.5 m from a patient who has received
350MBq of 18F.
High equivalent dose to the fingers can be received in preparation and
administration of radiopharmaceuticals, even if proper shielding is used. Injecting
eight patients per day with 400MBq of 99mTc per patient has been reported to
give a mean and maximum equivalent dose to the fingers of 80 and 330 mSv/a,
respectively, even if syringe shields are used. Without shielding, the maximum
equivalent dose will be about 2500 mSv/a.
Higher risk of internal exposure due to contamination is associated with
radioactive spills, animal experiments, emergency surgery of a therapy patient
and autopsy of a therapy patient. However, traces of the radionuclides used in
a nuclear facility can be found almost everywhere, especially on door handles,
taps, some specific equipment and in the patients toilet. Some procedures, such
as ventilation scans, might also cause contamination of both personnel and
equipment. Whole body measurements of workers have revealed an equilibrium
internal contamination of up to 10 kBq of 99mTc, which will result in an effective
dose of ~0.05 mSv/a. Although this is a small fraction of the external exposure,
every precaution must be taken to avoid contamination of the facility.


Of special concern is contamination of the skin, since this can result in

extremely high local equivalent doses. For instance, 1 kBq of 18F will result
in an initial equivalent dose rate to the skin of 0.8 mSv/h. The activity on the
hands after elution, preparation and administration of 99mTc radiopharmaceuticals
has been reported to be 0.02200 kBq, which results in an initial skin dose of
0.00550 mSv/h.
3.5.2. Justification, optimization and dose limitation
Nuclear medicine workers have no personal benefit from exposure.
Therefore, justification of occupational exposure must be included in justification
of the nuclear medicine practice itself. The risks in radiation work should not be
greater than for any other similar work. The upper limit of a tolerable risk for the
individual is determined by the dose limits (seeTable3.1). However, through
optimized protection, the incurred effective dose should be further reduced.
Besides facility and equipment design, shielding of sources, handling of sources
as well as personal protective equipment are important in the optimization
of occupational radiation protection. Optimization is also achieved through
education and training, resulting in awareness and involvement in radiation
From the examples above, it should be clear that the dose limits for
workers can be exceeded if the necessary protective precautions are not taken.
Radiation protection measures must be applied in each step of the work with
radiopharmaceuticals in the nuclear medicine facility, including work with the
The principal parties responsible for occupational exposure are licensees
and employers, and they should ensure that the exposure is limited and that
protection is optimized. The worker also has responsibilities and must follow
the rules and procedures as well as using the devices for monitoring and the
protective equipment and tools provided, and in all aspects cooperate with the
employer in order to improve the protection standard in the workplace.
3.5.3. Conditions for pregnant workers and young persons
It is generally accepted that the unborn child should be afforded the same
protection level as a member of the general public, meaning that a dose limit
of 1 mSv should be applied once pregnancy is declared. Good operational
procedures should ensure that the radiation doses received by staff working in
nuclear medicine facilities are well below any occupational dose limits. Therefore,
there is generally no need for a pregnant member of staff to change her duties
based on the expected dose to the embryo or fetus. However, removal of pregnant


women from work in laboratories where large quantities of radionuclides are

prepared and administered, and from nursing teams responsible for patients who
have been treated with radionuclides should be considered. These staff members
could receive a dose to the embryo or fetus comparable with the public dose limit
over the period of the pregnancy. Since all doses should be reduced whenever
possible, some supervisors will consider it prudent to reassign pregnant staff to
non-radiation duties if this is possible. Many nuclear medicine facility managers
would also accept requests from women to be reassigned to other duties for
reasons beyond radiation protection. Previous personal monitoring results can
help guide any decisions, noting that the dose to the fetus from external radiation
is not likely to exceed 25% of the personal dosimeter measurement.
According to the BSS, no person under the age of 16 years is to be
subjected to occupational exposure, and no person under the age of 18 years is to
be allowed to work in a controlled area unless supervised and then only for the
purpose of training.
3.5.4. Protective clothing
Suitable personal protective clothing should be provided for the use of all
persons employed in work in controlled areas. The protective clothes should be
adequate to prevent any contamination of the body of the worker for whom it is
provided and should include gloves, laboratory coats, safety glasses and shoes or
overshoes, as well as caps and masks for aseptic work.
A question frequently asked is whether lead aprons are useful for nuclear
medicine work. Wearing a lead apron at all times will reduce the effective dose
by a factor of about two. It is, therefore, a matter of judgement whether this dose
reduction compensates for the effort of wearing an apron. In some hospitals, lead
aprons are used in the case of prolonged injections and high activity.
3.5.5. Safe working procedures
The safety of the work in nuclear medicine is based on facility design as
well as on the use of protective clothing and the use of protective equipment
and tools as discussed above. These measures together with working procedures
aimed tominimize external exposure, risk of contamination and spread of
contamination, will optimize protection of workers. Work with unsealed sources
should always be supported by written local rules.
In order tominimize the risk of contamination in handling
radiopharmaceuticals, clean operation conditions and good laboratory practice
should be adopted, and protective clothing used. The work area should be
kept tidy and free from articles not required for work. It should be monitored


periodically and be cleaned often enough to ensureminimal contamination.

No food or drink, cosmetic or smoking materials, crockery or cutlery should
be brought into an area where unsealed radioactive substances are used. They
should not be stored in a refrigerator used for unsealed radioactive substances.
Handkerchiefs should never be used in these areas.
All manipulation for preparation, dispensing and administration of
radioactive materials should be carried out in such a way that the spread
of contamination isminimized. That includes preparing and dispensing
radiopharmaceuticals over a drip tray covered with absorbing paper as well as
using absorbing compresses at administrations. Any spills of radioactive material
should be immediately covered with absorbent material to prevent the spread of
material. If the spill cannot be cleaned up immediately, it must be marked to warn
other personnel of its location. Decontamination of the area must begin as soon
as possible.
When wearing gloves which may be contaminated, unnecessary contact
with all other objects should be avoided. Gloves should be removed and disposed
of in the radioactive waste bin as soon as work with radioactive substances is
After finishing work with the potential for contamination, the protective
clothing should be removed and placed in an appropriate container. Hands should
be washed and monitored.
In order tominimize external exposure, the three fundamental measures
of protection should be applied: time, distance and shielding. As far as possible,
the time of exposure should be as short as possible. Of course, this is important
in work where high exposure rates can be expected, such as in the preparation of
radiopharmaceuticals. However, limiting exposure time should not compromise
the quality of work or the use of other protective measures.
Direct handling of vials, syringes or other sources which produce a
significant radiation field is not recommended. Forceps or tongs should be used
to reduce the radiation exposure by increasing the distance between the source
and the hands. Properly designed vial and syringe shields must be used wherever
practicable. In cases where unshielded sources are handled or the exposure time
is prolonged, the work should be performed behind a properly designed lead
glass shield or similar type of protective barrier.
Radioactive waste should not be stored in the work area but transferred to a
separate radioactive waste storage room as soon as possible.
A patient undergoing a nuclear medicine imaging study is a source
of radiation exposure and contamination. Contact with these patients by
nursing staff presents little hazard, as the radiation dose rate is quite low, and
accumulated dose to any single individual would not be significant. However, for
nuclear medicine staff that spend a great deal of time in the immediate vicinity of


these patients, the accumulated radiation dose can be significant. These workers
should, whenever possible, maximize their distance from the patient and spend as
little time as possible in close proximity to the patient.
In summary, the following protective approaches can reduce external
exposure significantly:
For preparation and dispensing of radiopharmaceuticals, working behind a
lead glass bench shield, and using shielded vials and syringes;
For administration of radiopharmaceuticals to patients, using lead aprons in
the case of prolonged injection and high activity, and using a syringe shield;
During examinations, when the distance to the patient is short, using a
movable transparent shield.
3.5.6. Personal monitoring
The licensee and employer have the joint responsibility to ensure that
appropriate personal monitoring is provided to staff. This normally means that
the RPO would specify which workers need to be monitored routinely, the type
of monitoring device to be used and the body position where the monitor should
be worn, bearing inmind that some countries may have specific regulatory
requirements on these issues. Further, the regulatory body is likely to have
specified the monitoring period and the time frame for reporting monitoring
Staff to be monitored in a nuclear medicine facility should include all those
who work routinely with radionuclides or with the patients who have received
administrations of radiopharmaceuticals. This will include nursing staff who
either work routinely in nuclear medicine or nurse patients who have received
radionuclide therapy and staff dealing with excreta from radionuclide therapy.
Monitoring would not normally be extended to those that come into occasional
contact with nuclear medicine patients.
There are several types of external personal dosimetry systems and the
system to use is dependent on national or local conditions. In many countries,
the service is centralized to the regulatory body or provided through third party
personal dosimetry providers. Occasionally, some large hospitals have their own
personal dosimetry service. In all cases, the dosimetry provider must be approved
by the regulatory body.
Finger monitoring should be carried out occasionally on staff that regularly
prepare and administer radioactive substances to patients, and also when
setting up an operation which requires the routine handling of large quantities
of radionuclides. After handling unsealed radionuclides, the hands should be
monitored. It may, therefore, be convenient to mount a suitable contamination


monitor near the sink where hands are washed. Care should be taken to ensure
that the monitor itself does not become contaminated. In high background areas,
it will be necessary to shield the detector, and it may be convenient to have a foot
or elbow operated switch to activate the monitor.
Monitoring for internal contamination is rarely necessary in nuclear
medicine on radiation protection grounds but it may be useful in providing
reassurance to staff. The circumstances in which internal monitoring becomes
advisable are those where staff use significant quantities of 131I for thyroid therapy.
They should be included in a programme of thyroid uptake measurements.
In other circumstances where it is necessary to assess the intake of emitting
radionuclides (e.g. after a serious incident), the use of a whole body counter may
be appropriate. Such equipment should be available at national referral centres.
The possible use of an uncollimated gamma camera should also be considered.
Sometimes, a more detailed monitoring survey may be indicated if staff
doses have increased (or it is anticipated that they may do so in the future) as a
result of either the introduction of new examinations or procedures, or a change
in the nuclear medicine facilitys equipment. The RPO should decide who should
be monitored and at which monitoring sites.
Individual monitoring results must be analysed and records must be kept. It
is vital that the individual monitoring results are regularly assessed and the cause
of unusually high dosimeter readings should be investigated by the RPO, with
ensuing corrective actions where appropriate. The administrative arrangements,
the scope and nature of the individual monitoring records, and the length of time
for which records have to be kept may differ among countries.
3.5.7. Monitoring of the workplace
The BSS require licensees to develop programmes for monitoring the
workplace. Such programmes are described in Section 3.4.6 and in Chapters 9
and 20.
3.5.8. Health surveillance
According to the BSS, the licensee needs to make arrangements for
appropriate health surveillance in accordance with the rules established by
the national regulatory body. The primary purpose of health surveillance is to
assess the initial and continuing fitness of employees for their intended tasks.
The health surveillance programme should be based on the general principles of
occupational health.
No specific health surveillance related to exposure to ionizing radiation is
necessary for staff involved in nuclear medicine procedures. Only in the case of


overexposed workers at doses much higher than the dose limits would special
investigations involving biological dosimetry and further extended diagnosis and
medical treatment be necessary.
Counselling should be available to workers such as women who are
or may be pregnant, individual workers who have or may have been exposed
substantially in excess of dose limits and workers who may be worried about
their radiation exposure.
3.5.9. Local rules and supervision
According to the BSS, employers and licensees must, in consultation with
the workers or through their representatives:
Establish written local rules and procedures necessary to ensure adequate
levels of protection and safety for workers and other persons;
Include in the local rules and procedures the values of any relevant
investigation level or authorized level, and the procedure to be followed in
the event that any such value is exceeded;
Make the local rules and procedures, the protective measures and safety
provisions known to those workers to whom they apply and to other persons
who may be affected by them;
Ensure that any work involving occupational exposure be adequately
supervised and take all reasonable steps to ensure that the rules, procedures,
protective measures and safety provisions be observed.
These local rules should include all working procedures involving unsealed
sources in the facility such as:
Ordering radionuclides;
Unpacking and checking the shipment;
Storage of radionuclides;
General rules for work in controlled and supervised areas;
Preparation of radiopharmaceuticals;
Personal and workplace monitoring;
In-house transport of radionuclides;
Management of radioactive waste;
Administration of radiopharmaceuticals to the patients;
Protection issues in patient examinations and treatments;
Routine cleaning of facilities;
Decontamination procedures;
Care of radioactive patients.


It is the responsibilty of the licensee of the nuclear medicine facility to

ensure that local rules are established, maintained and continually reviewed. The
RPO would have significant involvement in this process.
3.6.1. Justification, optimization and dose limitation
According to the BSS, public exposure is exposure incurred by members of
the public from radiation sources, excluding any occupational or medical exposure.
The three ICRP principles described in Section 3.2.1 apply to public
exposure arising from the practice of nuclear medicine. Just as for occupational
exposure, the justification of public exposure is based on the justification of the
practice of nuclear medicine. The exposure of the general public is ultimately
restricted by the application of dose limits (seeTable3.1), but in the first instance
the application of the principle of optimization of protection ensures that public
doses will be ALARA.
The licensee is responsible for controlling public exposure arising from a
nuclear medicine facility. The presence of members of the public in or near the
nuclear medicine facility needs to be considered when designing the shielding
and flow of persons in the facility.
The sources of exposure of the general public are primarily the same as for
workers. Hence, the use of structural shielding and the control of sources, waste
and contamination are fundamental to controlling exposure of the public. There
are, however, some additional situations that need special consideration. These
include the release of patients examined or treated with radiopharmaceuticals.
3.6.2. Design considerations
The general layout of the nuclear medicine facility should take into account
the protection of members of the public. The areas for storage and preparation of
radiopharmaceuticals must be well separated from public areas such as waiting
rooms. The movement of radionuclides must beminimized. For example, the
room for preparation and dispensing of radiopharmaceuticals and the room for
administration should be adjacent and connected by a pass through. Areas where
significant activities of radionuclides are present must be appropriately shielded.
Access must be restricted so that members of the public are not allowed into
controlled areas. Radioactive waste must be stored in a secure location away from
areas accessible to the public. Since a patient still waiting for administration of
the radiopharmaceutical is regarded as a member of the public, separate waiting


rooms and toilets for injected and not injected patients should be considered in
order tominimize both external exposure and the spread of contamination.
3.6.3. Exposure from patients
Every precaution must be taken to ensure that the doses received by
individuals who come close to a patient or who spend some time in neighbouring
rooms remain below the dose limit for the public and below any applicable dose
constraint. For almost all diagnostic procedures, the maximum dose that could be
received by another person due to external exposure from the patient is a fraction
of the annual public dose limit and it should not normally be necessary to issue any
special radiation protection advice to the patient. One exception is restrictions on
breast-feeding a baby, which will be further discussed in Section Another
exception is an intensive use of positron emitters which may require structural
shielding based on the exposure of the public as discussed above (Section 3.4.4).
For patients who have undergone radionuclide therapy, specific advice should be
given regarding restrictions on their contact with other people. This is discussed
separately in Chapter20.
3.6.4. Transport of sources
One possible source of exposure of the general public is transport of
sources. It is performed both inside and outside the nuclear medicine facility.
Inside the facility, the transport includes distribution of the radioactive sources
from the storage area to where it will be used. Such transport should be limited
as far as possible by the facility design. The transport that takes place should
be performed according to optimized radiation protection conditions as given by
local rules.
The transport of radioactive sources to and from the nuclear medicine
facility should follow the internationally accepted IAEA Regulations for the Safe
Transport of Radioactive Material [3.8]. These Regulations include basic rules
for the transport itself and regulations about the shape and labelling of packages.
In general, the package is built in several parts. It should be mechanically
safe and reduce the effect of potential fire and water damage. The package should
be labelled with a sign. There are three different labels: IWhite, IIYellow and
IIIYellow. In all cases, the radionuclide and its activity should be specified. The
label gives some indication of the dose rate D at the surface of the package:
Category IWhite
D 0.005 mSv/h
Category IIYellow 0.005 < D 0.5 mSv/h
Category IIIYellow 0.5 < D 2 mSv/h


A more exact figure of the radiation around the package is given by the
transport index which is the maximum dose rate (mSv/h) at a distance 1 m from
the surface of the package multiplied by a hundred.
The detailed requirements given in Section 3 of the BSS are applicable to
medical exposure in nuclear medicine facilities. Furthermore, Ref.[3.9] describes
strategies to involve organizations outside the regulatory framework, such as
professional bodies (nuclear medicine physicians, medical physicists, nuclear
medicine technologists, radiopharmacists), whose cooperation is essential to
ensure compliance with the BSS requirements for medical exposures. Examples
that may illustrate this point include the adoption of protocols for calibration
of unsealed sources and for QA and for reporting accidental medical exposure.
Reference [3.3] provides further specific advice. A summary of the most relevant
issues for nuclear medicine is given in this section.
3.7.1. Justification of medical exposure
The BSS state that:
Medical exposures shall be justified by weighing the expected diagnostic
or therapeutic benefitsthat they yield against the radiation detriment
that they might cause, with account taken of the benefits and the risks of
available alternative techniques that do not involve medical exposure.
The principle of justification of medical exposure should not only be applied
to nuclear medicine practice in general but also on a case by case basis, meaning
that any examination should be based upon a correct assessment of the indications
for the examination, the actual clinical situation, the expected diagnostic and
therapeutic yields, and the way in which the results are likely to influence the
diagnosis and the medical care of the patient. The nuclear medicine specialist
has the ultimate responsibility for the control of all aspects of the conduct and
extent of nuclear medicine examinations, including the justification of the given
procedure for a patient. The nuclear medicine specialist should advise and make
decisions on the appropriateness of examinations and determine the techniques
to be used. In justifying a given diagnostic nuclear medicine procedure, relevant
international or national guidelines should be taken into account.
Any nuclear medicine procedure that occurs as part of a biomedical
research project (typically as a tool to quantify changes in a given parameter


under investigation) is considered justified if the project has been approved by an

ethics committee.
3.7.2. Optimization of protection
The principle of optimization of protection is applied to nuclear medicine
procedures that have been justified, and can be summarized as follows. For
diagnostic nuclear medicine procedures, the patient exposure should be
theminimum necessary to achieve the clinical purpose of the procedure,
taking into account relevant norms of acceptable image quality established by
appropriate professional bodies and relevant diagnostic reference levels (DRLs).
For therapeutic nuclear medicine procedures, the appropriate
radiopharmaceutical and activity are selected and administered so that the activity
is primarily localized in the organ(s) of interest, while the activity in the rest of
the body is kept ALARA.
The implementation of optimization of protection for patients in
nuclear medicine is quite complex and includes equipment design, choice of
radiopharmaceutical and activity, procedure considerations, DRLs, calibration,
clinical dosimetry and QA, as well as special considerations for children,
pregnant women and lactating women. This is further discussed in the following
sections. Administered activity and radiopharmaceuticals
For diagnostic procedures, it is necessary for the nuclear medicine specialist
in cooperation with the medical physicist to determine the optimum activity to
administer in a certain type of examination, taking the relevant DRL (seebelow)
into account. For any given procedure used on an individual patient, the optimum
activity will depend on the body build and weight of the patient, the patients
metabolic characteristics and clinical condition, the type of equipment used, the
type of study (static, dynamic, tomographic) and the examination time.
For a given type of imaging equipment, the diagnostic value of the
information obtained from an examination will vary with the amount of
administered activity. There is a threshold of administered activity below which
no useful information can be expected. Above this level, the diagnostic quality
will increase steeply with increasing activity. Once an acceptable image quality
has been reached, a further increase of the administered activity will only increase
the absorbed dose and not the value of the diagnostic information.
It should also be noted that limiting the administered activity below the
optimum, even for well intentioned reasons, will usually lead to a poor quality
of the result which may cause serious diagnostic errors. It is very important to


avoid failure to obtain the required diagnostic information; failure would result
in unnecessary (and, therefore, unjustified) irradiation and may also necessitate
repetition of the test.
If more than one radiopharmaceutical can be used for a procedure,
consideration should be given to the physical, chemical and biological properties
for each radiopharmaceutical, so as tominimize the absorbed dose and other risks
to the patient while at the same time providing the desired diagnostic information.
Other factors affecting the choice include availability, shelf life, instrumentation
and relative cost. It is also important that the radiopharmaceuticals used are
received from approved manufacturers and distributors, and are produced
according to national and international requirements. This is a requirement also
for in-house production of radiopharmaceuticals for PET studies.
The activity administered to a patient should always be determined and
recorded. Knowing the administered activity makes it possible to estimate the
absorbed dose to different organs as well as the effective dose to the patient.
Substantial reduction in absorbed dose from radiopharmaceuticals can be
achieved by simple measures such as hydration of the patient, use of thyroid
blocking agents and laxatives. Optimization of protection in procedures
The nuclear medicine procedure starts with the request for an examination or
treatment. The request should be written and contain basic information about the
patients condition. This information should help the nuclear medicine specialist
to decide about the most appropriate method to use and to decide how urgent
the examination is. The patient should then be scheduled for the examination or
treatment and be informed about when and where it will take place. Some basic
information about the procedure should also be given, especially if it requires
some preparation of the patient, such as fasting. These initial measures require
an efficient and reliable administrative system. In parallel to these routines, the
nuclear medicine facility has to ensure that the radiopharmaceutical to be used is
available at the time of the scheduled procedure.
When the patient appears in the nuclear medicine facility, they should
be correctly identified using the normal hospital or clinic routines. The patient
should be informed about the procedure and have the opportunity to ask questions
about it. A fully informed and motivated patient is the basis for a successful
examination or treatment. Before the administration of the radiopharmaceutical,
the patient should be interviewed about possible pregnancy, small children at
home, breast-feeding and other relevant questions which might have implications
for the procedure. Before administration, the technologist or doctor should
check the request and ensure that the right examination or treatment is scheduled


and that the right radiopharmaceutical and the right activity are dispensed. If
everything is in order, the administration can proceed. The administered activity
should always be recorded for each patient.
While most adults can maintain a required position without restraint
or sedation during nuclear medicine examinations, it may be necessary to
immobilize or sedate children, so that the examination can be completed
successfully. Increasing the administered activity to reduce the examination time
is an alternative that can be used in elderly patients with pain.
Optimization of protection in an examination means that equipment should
be operated within the conditions established in the technical specifications, thus
ensuring that it will operate satisfactorily at all times, in terms of both the tasks to
be accomplished and radiation safety. More details are given in Chapters 8 and 15.
Particular procedural considerations for children, pregnant women and lactating
women are given in the following subsections.
Optimization of protection in radionuclide therapy means that a correctly
calculated and measured activity should be administered to the patient in order
to achieve the prescribed absorbed dose in the organ(s) of interest, while the
radioactivity in the rest of the body is kept as low as reasonably achievable.
Optimization also means using routines to avoid accidental exposures of the
patient, the staff and members of the general public. Radionuclide therapy is
further discussed in Chapter20.
The availability of a written manual of all procedures carried out by the
facility is highly desirable. The manual should regularly be revised as part of a
QA programme. Pregnant women
Special consideration should be given to pregnant women exposed to
ionizing radiation due to the larger probability of inducing radiation effects in
individuals exposed in utero compared to exposed adults. As a basic rule, it is
recommended that diagnostic and therapeutic nuclear medicine procedures
of women likely to be pregnant be avoided unless there are strong clinical
In order to avoid unintentional irradiation of the unborn child, a female of
childbearing age should be evaluated regarding possible pregnancy or a missed
period. This should be done when interviewing and informing the woman prior to
the examination or treatment. It is also common to place a poster in the waiting
area requesting a woman to notify the staff if she is or thinks she is pregnant.
If the patient is found not to be pregnant without any doubt, the examination
or treatment can be performed as planned. If pregnancy is confirmed,
careful consideration should be given to other methods of diagnosis or to the


postponement of the examination until after delivery. If, after consultation

between the referring physician and the nuclear medicine specialist, these options
are not feasible, then the examination should be performed, but the process of
optimization of protection needs to also consider protection of the embryo/fetus.
In order to reduce the fetal dose, it may sometimes be possible to reduce
the administered activity and acquire images for longer times, but great care must
be taken not to compromise the quality of the result. After the administration of
radiopharmaceuticals, frequent voiding should be ensured tominimize exposure
from the bladder. This contribution to the fetal dose can be further reduced by
administering the radiopharmaceutical when the bladder is partially filled, rather
than immediately after voiding.
Of special concern is also the use of CT in PET/CT or SPECT/CT
examinations. Routine diagnostic CT examinations of the pelvic region with and
without contrast injection can lead to a dose of 50 mSv to the uterus which is
assumed to be equivalent to the fetal dose in early pregnancy. It is important to
use low dose CT protocols and to reduce the scanning area to aminimum when
PET/CT or SPECT/CT scanning is indicated in a pregnant patient.
Pregnant women should not be subject to therapy with a radioactive
substance unless the application is life-saving. Following treatment with a
therapeutic activity of a radionuclide, female patients should be advised to avoid
pregnancy for an appropriate period. More details are given in Ref.[3.3].
If the fetal dose is suspected to be high (e.g. >10 mSv), it should be carefully
determined by a qualified medical physicist and the pregnant woman should be
informed about the possible risks. The same procedure should be applied in the
case of an inadvertent exposure, which can be incurred by a woman who later
was found to have been pregnant at the time of the exposure or in emergency
Exposure of a pregnant patient at a time when the pregnancy was not known
often leads to her apprehension because of concern about the possible effects on
the fetus. It may lead to a discussion regarding termination of pregnancy due to
the radiation risks. Many misunderstandings and lack of knowledge, also among
physicians, have probably resulted in unnecessary termination of pregnancies.
It is generally considered that for a fetal dose of less than 100 mGy, as in most
diagnostic procedures, termination of pregnancy is not justified from the point
of radiation risks. At higher doses, individual circumstances should be taken into
account. This is an ethical issue and the national authorities should give guidance. Lactating women
When nuclear medicine examinations are requested for women who are
breast-feeding, they present a potential radiation hazard to the baby. This is due


to uptake of some radiopharmaceuticals in breast tissue followed by excretion

into the breast milk. The dose to the baby depends on various factors such as the
radiopharmaceutical, the amount of milk and the time between the administration
of the radiopharmaceutical to the mother and the feeding of the child. The mother
also represents a source of external exposure and contamination when feeding
or cuddling the baby. The dose will depend on the time the child is held, the
distance from the mothers body and personal hygiene. Some restrictions on
breast-feeding and advice to the mother are necessary in order tominimize the
exposure of the baby to an acceptable level. The baby is a member of the public
and a typical constraint on the dose from a single source of exposure (in this case,
per episode) is 0.3 mSv.
Before a nuclear medicine examination or therapy with radionuclides,
the woman should be asked, orally or in writing, whether she is breast-feeding
a child. A notice requesting the patient to inform the staff about breast-feeding
should also be prominently displayed in the waiting area. If the answer is yes,
consideration should be given as to whether the examination or treatment could
reasonably be delayed until she has ceased breast-feeding. If not, advice about
restriction of breast-feeding dependent on the diagnostic or therapeutic procedure
should be given to the patient.
It is the responsibility of the nuclear medicine specialist in cooperation with
the medical physicist to establish local rules regarding breast-feeding and close
contact between the mother and the child after a nuclear medicine examination or
treatment. The rules should be based on recommendations given by international
and national authorities as well as professional organizations. Some guidance is
found in Ref.[3.3]. Children
Optimization of protection for an examination of a child is basically an
optimization of the administered activity. There are several approaches to the
problem of how to calculate the administered activity for children. It should be
theminimum consistent with obtaining a diagnostic result. As this is the same
principle which is applied to adult doses, the normal activity administered to
adults should be used as a guide, bearing inmind that the average adult body
weight is 70 kg. For children or young persons, body weight should always
be measured and the adult administered activity should then be scaled down.
Opinions differ as to how the scaling should be achieved. Simply reducing the
activity in proportion to body weight may, in some types of investigation, result
in inadequate image quality. Another method is based on the principle of scaling
in proportion to body surface area. This approach should give the same image
count density as that for an adult patient, although the effective dose is higher. As


a general guide, activities less than 10% of the normal adult activity should not
be administered.
In hybrid imaging, the CT protocol should be optimized by reducing the
tube currenttime product (mAs) and tube potential (kV) without compromising
the diagnostic quality of the images. Careful selection of slice width and pitch as
well as scanning area should also be done. It is important that individual protocols
based on the size of the child are used. The principles behind such protocols
should be worked out by the medical physicist and the responsible specialist.
Since the examination times in nuclear medicine examinations are quite
long, there may be problems in keeping the child still during the examination.
Even small body motions can severely interfere with the quality of the
examination and make it useless. There are several methods of mechanical
support to fasten the child. Drawing the childs attention to something else such
as a television programme can also be useful for older children. Sometimes, even
sedation or general anaesthesia may be necessary. Calibration
The licensee of a nuclear medicine facility needs to ensure that a dose
calibrator or activity meter is available for measuring activity in syringes or vials.
The validity of measurements should be ensured by regular quality control of
the instrument, including periodic reassessment of its calibration, traceable to
secondary standards. Clinical (patient) dosimetry
The licensee of a nuclear medicine facility should ensure that appropriate
clinical dosimetry by a medical physicist is performed and documented. For
diagnostic nuclear medicine, this should include representative typical patient
doses for common procedures. For therapeutic nuclear medicine, this needs to
be for each individual patient, and includes absorbed doses to relevant organs or
tissues. Diagnostic reference levels
Many investigations have shown a large spread of administered activities
for a certain type of diagnostic nuclear medicine examination between different
hospitals within a country, even if the equipment used is similar in performance.
Even though no dose limits are applied to medical exposure, the process of
optimization should result in about the same administered activity for the same
type of examination and for the same size of patient.


The concept of a DRL provides a tool for the optimization of protection

in medical exposure. In the case of nuclear medicine, the DRL is given as
administered activity for a certain type of examination and for a normal sized
patient. DRLs are aimed to assist in the optimization of protection by helping to
avoid unnecessarily high activities to the patient or too low activities to provide
useful diagnostic information. DRLs are normally set at the national level as a
result of consultation between the health authority, relevant professional bodies
and the radiation protection regulatory body. Quality assurance for medical exposures
The BSS require the licensee of the nuclear medicine facility to have a
comprehensive programme of QA for medical exposures. The programme needs
to have the active participation of the medical physicists, nuclear medicine
specialists, nuclear medicine technologists and radiopharmacists, and needs to
take into account principles established by international organizations, such as
the WHO and PAHO, and relevant professional bodies.
The programme of QA for medical exposures should be complementary to
and part of the wider programme of QA for radiation protection the latter also
including occupational and public exposure. In turn, this programme needs to be
part of and harmonized with the nuclear medicine facilitys quality management
system. Section 3.9 discusses the wider QA programme, while the remainder of
this subsection deals with some aspects of the programme as it applies to medical
The programme of QA for medical exposures should include:
Measurements by, or under the oversight of, a medical physicist of the
physical parameters of medical radiological equipment at the time of
acceptance and commissioning prior to clinical use on patients, periodically
thereafter, and after any major maintenance that could affect patient
Implementation of corrective actions if measured values of the physical
parameters are outside established tolerance limits;
Verification of the appropriate physical and clinical factors used in patient
diagnosis or treatment;
Records of relevant procedures and results;
Periodic checks of the appropriate calibration and conditions of operation
of dosimetry and monitoring of equipment.
In addition, the licensee needs to ensure that there are regular and
independent audits of the programme of QA for medical exposures, their


frequency depending on the complexity of the nuclear medicine procedures

performed and the risks involved.
The above indicates, among other actions, the need for quality control tests
on the equipment. More details regarding quality control of equipment used in
diagnosis will be found in other chapters of this book.
3.7.3. Helping in the care, support or comfort of patients
Certain patients, such as children, the elderly or the infirm, may have
difficulty during a nuclear medicine procedure. Occasionally, people knowingly
and voluntarily (other than in their employment or occupation) may volunteer
to help in the care, support or comfort of patients. In such circumstances, the
dose to these persons (excluding children and infants) should be constrained
so that it is unlikely that it will exceed 5 mSv during the period of a patients
diagnostic examination or treatment. The dose to children visiting patients who
have ingested radioactive materials should be similarly constrained to less than
1 mSv. Special concern should be given to members of the family of a patient
who has received radionuclide therapy. This is further discussed in Chapter20.
Sometimes, a nurse escorting a patient to the nuclear medicine facility is
asked to provide assistance during a procedure. Any resultant exposure should
be regarded as occupational, and the nurse should have received education and
training on this role.
3.7.4. Biomedical research
The exposure of humans for biomedical research is deemed not to be
justified unless it is in accordance with the provisions of the Helsinki Declaration
[3.10] and follows the guidelines for its application prepared by the Council for
International Organizations of Medical Sciences [3.11]. It is also subject to the
approval of an ethics committee.
The use of radioactive trace substances is common in biomedical research.
Diagnostic nuclear medicine procedures may be part of a biomedical research
project, typically as a means for quantifying changes in a given parameter
under investigation or assessing the efficacy of a treatment under investigation.
An exposure as part of biomedical research is treated on the same basis as a
medical exposure and, therefore, is not subject to dose limits. However, in all
investigations involving exposure of humans, a careful estimation of the radiation
dose to the volunteer should be made. The associated risk should then be weighed
against the benefit for the patient or society. Recommendations are given by the
ICRP. The BSS require the use of dose constraints, on a case by case basis, in the
process of optimization.


3.7.5. Local rules

The management of patients in the nuclear medicine facility should be
supported by written local rules covering all procedures that may affect medical
exposure. These local rules should be signed by the responsible person and
known to every member of the staff and should include:
Routines for patient identification and information;
Prescribed radiopharmaceutical and activity for adults and children for
different types of examination, including methods used to adjust the activity
to the single patient and routes of administration;
Management of patients that are pregnant or might be pregnant;
Management of breast-feeding patients;
Routines for safe preparation and administration of radiopharmaceuticals
including activity measurements;
Procedures in case of misadministration of the radiopharmaceutical;
Detailed procedure manuals for every type of examination including
handling of equipment.
3.8.1. Safety assessment and accident prevention
Unintended and accidental exposure may occur due to equipment failure,
human error or a combination of both. Although such events can be identified
by a careful safety assessment, their details and the time of occurrence cannot be
predicted. These exposures are called potential exposures. It is the responsibility
of the licensee to take measures in order to prevent such events as far as possible
and, in case they occur, mitigate their consequences.
According to the BSS, the licensee needs to conduct a safety assessment
applied to all stages of the design and operation of the nuclear medicine facility,
and present the report to the regulatory body if required. The safety assessment
needs to include, as appropriate, a systematic critical review of identification of
possible events leading to unintended or accidental exposure. In practice, this
means that all procedures in which unsealed sources are involved in the work
should be listed and for every procedure it should be asked what can go wrong.
Some examples are given in Table3.2.




Procedure and involvement

What can go wrong?

Patients involved
Request and scheduling

Wrong patient scheduled

Identification at arrival

Wrong patient identified


Missed pregnancy or breast-feeding

Administration of

Misadministration (wrong patient, wrong

activity, wrong radiopharmaceutical)


Contamination of waiting area (vomiting,



Inconclusive due to contamination,

equipment and/or software failure

Workers involved
Ordering of sources

Unauthorized ordering

Receipt and unpacking of shipments

Damage to package, contamination

Storage of sources

Unshielded sources, high dose rates,

loss of sources

Preparation and administration of


High doses recorded, contamination of

workers and facilities

Handling of radioactive waste

Contamination of workers and facilities

General public involved

Storage of sources

Loss of sources

Handling of sources

Contamination of facility

Radioactive waste

Loss of sources, contamination of facilities

Radioactive patient

Escape of hospitalized patient, medical

emergency, death of patient

Undertaking a safety assessment requires using ones imagination to try to

define an event that could result in a potential exposure, even if the event has
never occurred before. For instance, what should be done if a patient who just
received 15 GBq of 131I escapes from the isolation ward and the hospital and is
seriously injured in a road accident?
If an unintended or accidental medical exposure occurs, the licensee is
required to determine the patient doses involved, identify any corrective actions


needed to prevent recurrence and implement the corrective measures. There may
be a requirement to report the event to the regulatory body.
A well established RPP is fundamental in accident prevention together with
a high level of safety culture in the organization and among the people working
in a nuclear medicine facility. The content of an RPP as well as the importance of
well established working procedures in order to protect patients, workers and the
general public have been discussed in the sections above. It should be stressed
that documentation of the procedures used in the facility is also important in
accident prevention. Other important factors are a well working QA programme
and a programme for continuing education and training which includes not only
the normal practices, but also accidental situations and lessons learned from
3.8.2. Emergency plans
According to the BSS, the licensee needs to prepare emergency procedures
on the basis of events identified by the safety assessment. The procedures should
be clear, concise and unambiguous, and need to be posted visibly in places where
their need is anticipated. An emergency plan needs to, as aminimum, list and
Predictable incidents and accidents, and measures to deal with them;
The persons responsible for taking actions, with full contact details;
The responsibilities of individual personnel in emergency procedures
(nuclear medicine physicians, medical physicists, nuclear medicine
technologists, etc.);
Equipment and tools necessary to carry out the emergency procedures;
Training and periodic drills;
The recording and reporting system;
Immediate measures to avoid unnecessary radiation doses to patients, staff
and the public;
Measures to prevent access of persons to the affected area;
Measures to prevent spread of contamination.
The most likely accident in a nuclear medicine facility is contamination
of workers, patients, equipment and facilities. It can range from small to very
large spillages of radioactivity, for example, serious damage to the technetium
generator or spillage of several gigabecquerels of 131I. The procedures of cleaning
up a small amount of contamination should be known and practised by every
technologist in the facility. The cleaning procedures in cases of more severe
contamination should always be supervised by the RPO. Local rules should be


established that define serious contamination based on radionuclide, activity

and whether it is contamination of a person or equipment and facilities. It is
recommended that the facility have an emergency kit readily available in case of
contamination. Such a kit should contain:
Protective clothing, e.g. overshoes, gloves;
Decontamination materials for the affected areas, including absorbent
materials for wiping up spills;
Decontamination materials for persons;
Warning notices;
Portable monitoring equipment (in working order and regularly checked);
Bags for waste, tape, labels, pencils.
Several severe accidents in medical exposures in nuclear medicine
have been reported and are solely associated with radionuclide therapy and
especially when using 131I in treatment of thyroid diseases. Several incidents
with misadministration of radiopharmaceuticals in diagnostic nuclear medicine
have also been reported. These include examination of the wrong patient or
administration of the wrong radiopharmaceutical or the wrong activity. The most
common incident is to administer the wrong radiopharmaceutical. Even if this
does not cause severe injury to the patient, it is a non-justified exposure with
increased radiation risks. It will also lead to a delayed diagnosis, increased cost
and increased workload because the examination will have to be repeated. Last
but not least, it will cause reduced confidence in the practice of nuclear medicine.
Other accidents and incidents that also involve the general public include
the possible death of a patient containing radionuclides. In diagnostic nuclear
medicine, such an incident can generally be left without specific measures.
However, in radionuclide therapy, emergency plans have to be available on
how to handle the cadaver. Since this is a sensitive issue, depending on ethical
and religious rules and traditions, advice should be available from the national
3.8.3. Reporting and lessons learned
In the event of an incident or accident, the licensee has the responsibility
to ensure that a comprehensive investigation takes place and a report is produced
that includes the following information:
A description of the incident by all persons involved;



Methods used to estimate the radiation dose received by those involved

in the incident and implications of those methods for possible subsequent
Methods used to analyse the incident and to derive risk estimates from the
The subsequent medical consequences for those exposed;
The particulars of any subsequent legal proceedings that may ensue;
Conclusions drawn from the evaluation of the incident and recommendations
on how to prevent a recurrence of such an accident.
In the case of a misadministration or an accident in radionuclide therapy,
the responsible nuclear medicine specialist should be promptly informed. They
should then inform the referring physician and the patient. The medical physicist
should make dose calculations and the staff involved in the accident should
independently describe their view of the accident. Conclusions regarding any
deficits in the procedures should be drawn and necessary changes implemented.
Finally, the licensee may need to submit the report to the regulatory body.
In order to avoid future accidents, it is important to learn from previous
ones. The initiating event and the contributing factors can always be identified.
This information provides material that should be used to prevent future
accidents. This is achieved by informing all members of staff about the accident
or incident, which means that it is very important to have an efficient reporting
system and a programme for local education and training that also includes
potential exposures.
3.9.1. General considerations
The International Organization for Standardization defines QA as all
planned and systematic actions needed to provide confidence that a structure,
system or component will perform satisfactorily in service. Satisfactory
performance in nuclear medicine implies the optimum quality of the entire
process. Since an examination or therapy is justified only if the procedure
benefits the patient, QA in the whole process of nuclear medicine is an important
aspect of radiation protection.
The BSS require the licensee of the nuclear medicine facility to have
established a QA programme that provides adequate assurance that the specified
requirements relating to protection and safety are satisfied, and that provides



quality control mechanisms and procedures for reviewing and assessing the
overall effectiveness of protection and safety measures.
It is a common and growing practice that hospitals or clinics implement a
quality management system for all of the medical care received in diagnosis and
treatment, i.e. covering the overall nuclear medicine practice. The QA programme
envisaged by the BSS should be part of the wider facility quality management
system. In the hospital or clinic, it is common to include QA as part of the RPP
or, conversely, to include the RPP as part of a more general QA programme for
the hospital or clinic. Regardless of its organization, it is important that radiation
protection is an integral part of a system of quality management. The remainder
of this section considers aspects of QA applied to a nuclear medicine facility that
are covered in the BSS. Specific details with respect to medical exposure are
covered in Section
An effective QA programme requires a strong commitment from the
nuclear medicine facilitys management to provide the necessary resources of
time, personnel and budget. It is recommended that the nuclear medicine facility
establish a group that actively works with QA issues. Such a QA committee
should have a representative from management, a nuclear medicine physician, a
medical physicist, a nuclear medicine technologist and an engineer as members.
The QA committee should meet regularly and review the different components of
the programme.
The QA programme should cover the entire process from the initial decision
to adopt a particular procedure through to the interpretation and recording
of results, and should include ongoing auditing, both internal and external, as
a systematic control methodology. The maintenance of records is an important
part of QA. One important aspect of any QA programme is continuous quality
improvement. This implies a commitment of the staff to strive for continuous
improvement in the use of unsealed sources in diagnosis and therapy, based
on new information learned from their QA programme and new techniques
developed by the nuclear medicine community at large. Feedback from
operational experience and lessons learned from accidents or near misses can
help identify potential problems and correct deficiencies, and should, therefore,
be used systematically, as part of the continuous quality improvement.
A QA programme should cover all aspects of diagnosis and therapy,
The prescription of the procedure by the medical practitioner and its
documentation (supervising if there is any error or contraindication);
Appointments and patient information;
Clinical dosimetry;
Optimization of examination protocol;


Record keeping and report writing;

Quality control of radiopharmaceuticals and radionuclide generators;
Acceptance and commissioning;
Quality control of equipment and software;
Waste management procedures;
Training and continuing education of staff;
Clinical audit;
General outcome of the nuclear medicine service.
Further details on the general components of a QA programme and
the associated quality control tests are given in Ref.[3.3]. The WHO has also
published guidelines on QA in nuclear medicine [3.12], covering the organization
of services, the training of personnel, the selection of procedures, quality control
requirements for instrumentation and radiopharmaceuticals, as well as the
interpretation and evaluation of results. The IAEA has several other relevant
publications on QA for various aspects of nuclear medicine (seethe Bibliography
for details).
3.9.2. Audit
The QA programme should be assessed on a regular basis either as an
external or internal audit or review. Audits of activities within the QA programme
should be scheduled on the basis of the status and importance of the activity.
Management should establish a process for such assessments to identify and
correct administrative and management problems that may prevent achievement
of the objectives. Audits and reviews should be conducted by persons who are
technically competent to evaluate the processes and procedures being assessed,
but do not have any direct responsibility for those activities. These may be staff
from other work areas within the organization (internal audit), or an independent
assessment by other organizations (external audit). External audits are generally
a requirement for an accredited practice.
The quality audit should be performed in accordance with written
procedures and checklists. It should include medical, technical and procedural
checks, with the objective to enhance the effectiveness and efficiency of the QA
programme. Any major changes in the QA programme should initiate an audit.
The result should be documented and necessary correction initiated and followed



Recommendations of the ICRP, Publication 103, Elsevier (2008).



Protection and Safety of Radiation Sources: International Basic Safety Standards,
IAEA Safety Standards Series No. GSR Part 3, IAEA, Vienna (2014).
Standards in Nuclear Medicine, Safety Reports Series No. 40, IAEA, Vienna (2005).
[3.4] INTERNATIONAL ATOMIC ENERGY AGENCY, Nuclear Medicine Resources
Manual, IAEA, Vienna (2006).
Protection, IAEA Safety Standards Series No. RS-G-1.1, IAEA, Vienna (1999).
[3.6] INTERNATIONAL ATOMIC ENERGY AGENCY, Assessment of Occupational
Exposure Due to Intakes of Radionuclides, IAEA Safety Standards Series
No. RS-G-1.2, IAEA, Vienna (1999).
[3.7] INTERNATIONAL ATOMIC ENERGY AGENCY, Assessment of Occupational
Exposure Due to External Sources of Radiation, IAEA Safety Standards Series
No. RS-G-1.3, IAEA, Vienna (1999).
[3.8] INTERNATIONAL ATOMIC ENERGY AGENCY, Management of Waste from
the Use of Radioactive Material in Medicine, Industry, Agriculture, Research and
Education, IAEA Safety Standards Series No. WS-G-2.7, IAEA, Vienna (2005).
[3.9] INTERNATIONAL ATOMIC ENERGY AGENCY, Regulations for the Safe Transport
of Radioactive Material, 2012 Edition, IAEA Safety Standards Series No. SSR-6,
IAEA, Vienna (2012).
[3.10] WORLD MEDICAL ASSOCIATION, 18th World Medical Assembly, Helsinki, 1974,
as amended by the 59th World Medical Assembly, Seoul (2008).
WORLD HEALTH ORGANIZATION, International Ethical Guidelines for
Biomedical Research Involving Human Subjects, CIOMS, Geneva (2002).
[3.12] WORLD HEALTH ORGANIZATION, Quality Assurance in Nuclear Medicine,
WHO, Geneva (1982).



EUROPEAN COMMISSION, European Guidelines on Quality Criteria for Computed
Tomography, Rep. EUR 16262 EN, Office for Official Publications of the European
Communities, Brussels (1999).
Quality Control of Nuclear Medicine Instruments 1991, IAEA-TECDOC-602 (1991).
Radiological Protection for Medical Exposure to Ionizing Radiation, IAEA Safety Standards
Series No. RS-G-1.5 (2002).
IAEA Quality Control Atlas for Scintillation Camera Systems (2003).
Quality Assurance for Radioactivity Measurement in Nuclear Medicine, Technical Reports
Series No. 454 (2006).
Radiation Protection in Newer Medical Imaging Techniques: PET/CT, Safety Reports Series
No. 58 (2008).
Quality Assurance for PET and PET/CT Systems, IAEA Human Health Series No. 1 (2009).
Quality Assurance for SPECT Systems, IAEA Human Health Series No. 6 (2009).
Quality Management Audits in Nuclear Medicine Practices (2009).
Radiation Protection of Patients (RPoP),
Radiological Protection of the Worker in Medicine and Dentistry, Publication 57, Pergamon
Press, Oxford and New York (1989).
Radiological Protection in Biomedical Research, Publication 62, Pergamon Press, Oxford and
New York (1991).
Radiological Protection in Medicine, Publication 105, Elsevier (2008).
Pregnancy and Medical Radiation, Publication 84, Pergamon Press, Oxford and New York
Quantities and Units in Radiation Protection Dosimetry, ICRU Rep. 51, Bethesda MD (1993).
MADSEN, M.T., et al., AAPM Task Group 108: PET and PET/CT shielding requirements,
Med. Phys. 33 (2006) 1.
Quality Standards in Nuclear Medicine, IPSM Rep. No. 65, York (1992).


Department of Radiology, Oncology and Radiation Science,
Uppsala University,
Uppsala, Sweden
All matter in the universe has its origin in an event called the big bang,
a cosmic explosion releasing an enormous amount of energy about 14 billion
years ago. Scientists believe that particles such as protons and neutrons, which
form the building blocks of nuclei, were condensed as free particles during the
first seconds. With the decreasing temperature of the expanding universe, the
formation of particle combinations such as deuterium (heavy hydrogen) and
helium occurred. For several hundred million years, the universe was plasma
composed of hydrogen, deuterium, helium ions and free electrons. As the
temperature continued to decrease, the electrons were able to attach to ions,
forming neutral atoms and converting the plasma into a large cloud of hydrogen
and helium gas. Locally, this neutral gas slowly condensed under the force of
gravity to form the first stars. As the temperature and the density in the stars
increased, the probability of nuclear fusion resulting in the production of heavier
elements increased, culminating in all of the elements in the periodic table that
we know today. As the stars aged, consuming their hydrogen fuel, they eventually
exploded, spreading their contents of heavy materials around the universe.
Owing to gravity, other stars formed with planets around them, composed of
these heavy elements. Four and a half billion years have passed since the planet
Earth was formed. In that time, most of the atomic nuclei consisting of unstable
protonneutron combinations have undergone transformation (radioactive
decay) to more stable (non-radioactive) combinations. However, some with very
long half-lives remain: 40K, 204Pb, 232Th and the naturally occurring isotopes of
The discovery of these radioactive atoms was first made by Henri Becquerel
in 1896. The chemical purification and elucidation of some of the properties
of radioactive substances was further investigated by Marie Skodowska-Curie
and her husband Pierre Curie. Since some of these long lived radionuclides
generated more short lived radionuclides, a new scientific tool had been


discovered that was later found to have profound implications in what today is
known as nuclear medicine. George de Hevesy was a pioneer in demonstrating
the practical uses of the new radioactive elements. He and his colleagues used a
radioactive isotope of lead, 210Pb, as a tracer (or indicator) when they studied the
solubility of sparingly soluble lead salts. De Hevesy was also the first to apply
the radioactive tracer technique in biology when he investigated lead uptake in
plants (1923) using 212Pb. Only one year later, Blumengarten and Weiss carried
out the first clinical study, when they injected 212Bi into one arm of a patient and
measured the arrival time in the other arm. From this study, they concluded that
the arrival time was prolonged in patients with heart disease.
4.1.1. Induced radioactivity
In the beginning, nature was the supplier of the radioactive nuclides
used. Isotopes of uranium and thorium generated a variety of radioactive
heavy elements such as lead, but radioactive isotopes of light elements were
not known. Marie Curies daughter Irne, together with her husband Frdric
Joliot took the next step. Alpha emitting sources had long been used to bombard
different elements, for example, by Ernest Rutherford who studied the deflection
of particles in gold foils. The large deflections observed in this work led to
the conclusion that the atom consisted of a tiny nucleus of protons with orbiting
electrons (similar to planets around the sun). However, JoliotCurie also showed
that the particles induced radioactivity in the bombarded foil (in their case,
aluminium foil). The induced radioactivity had a half-life of about 3min. They
identified the emitted radiation to be from 30P created in the nuclear reaction
Al(, n)30P.
They also concluded that:
These elements and similar ones may possibly be formed in different
nuclear reactions with other bombarding particles: protons, deuterons
and neutrons. For example, 13N could perhaps be formed by capture of a
deuteron in 12C, followed by the emission of a neutron.
The same year, this was proved to be true by Ernest Lawrence in Berkeley,
California and Enrico Fermi in Rome. Lawrence had built a cyclotron capable
of accelerating deuterons up to about 3MeV. He soon reported the production
of 13N with a half-life of 10min. Thereafter, the cyclotron was used to produce
several other biologically important radionuclides such as 11C, 32P and 22Na.
Fermi realized that the neutron was advantageous for radionuclide production.
Since it has no charge, it could easily enter into the nucleus and induce a nuclear
reaction. He immediately made a strong neutron source by sealing up 232Ra gas


with beryllium powder in a glass vial. The particle emitted from 232Ra caused a
nuclear reaction in beryllium and a neutron was emitted, 9Be(, n)12C.
Fermi and his research group started a systematic search by irradiating all
available elements in the periodic system with fast and slow neutrons to study
the creation of induced radioactivity. From hydrogen to oxygen, no radioactivity
was observed in their targets, but in the ninth element, fluorine, their hopes were
fulfilled. In the following weeks, they bombarded some 60 elements and found
induced radioactivity in 40 of them. They also observed that the lighter elements
were usually transmuted into radionuclides of a different chemical element,
whereas heavier elements appeared to yield radioisotopes of the same element
as the target.
These new discoveries excited the scientific community. From having been
a rather limited technique, the radioactivity tracer principle could suddenly be
applied in a variety of fields, especially in life sciences. De Hevesy immediately
started to study the uptake and elimination of 32P phosphate in various tissues of
rats and demonstrated, for the first time, the kinetics of vital elements in living
creatures. Iodine-128 was soon after applied in the diagnosis of thyroid disease.
This was the start of the radiotracer technology in biology and medicine as
we know it today.
One early cyclotron produced nuclide of special importance was 11C
since carbon is fundamental in life sciences. Carbon-11 had a half-life of only
20min but by setting up a chemical laboratory close to the cyclotron, organic
compounds labelled with 11C were obtained in large amounts. Photosynthesis
was studied using 11CO2 and the fixation of carbon monoxide in humans by
inhaling 11CO. However, 20min was a short half-life and the use of 11C was
limited to the most rapid biochemical reactions. It must be remembered that the
radio-detectors used at that time were primitive and that the chemical, synthetic
and analytical tools were not adapted to such short times. A search to find a more
long lived isotope of carbon resulted in the discovery in 1939 of 14C produced in
the nuclear reaction 13C(d, p)14C.
Unfortunately, 14C produced this way was of limited use since the
radionuclide could not be separated from the target. However, during the
bombardments, a bottle of ammonium nitrate solution had been standing close
to the target. By pure chance, it was discovered that this bottle also contained
C, which had been produced in the reaction 14N(n, p)14C.
The deuterons used in the bombardment consist of one proton and one
neutron with a binding energy of about 2MeV. When high energy deuterons
hit a target, it is likely that the binding between the particles breaks and that a
free neutron is created in what is called a stripping reaction. The bottle with
ammonium nitrate had, thus, unintentionally been neutron irradiated. Since no
carbon was present in the bottle (except small amounts from solved airborne


carbon dioxide), the 14C produced this way was of high specific radioactivity.
It was also very easy to separate from the target. In the nuclear reaction, a
hot carbon atom was created, which formed 14CO2 in the solution. By simply
bubbling air through the bottle, the 14C was released from the target.
The same year, tritium was discovered by deuteron irradiation of water.
One of the pioneers Martin Kamen stated:
Within a few months, after the scientific world had somewhat ruefully
concluded that development of tracer techniques would be seriously
handicapped because useful radioactive tracers for carbon, hydrogen,
oxygen and nitrogen did not exist, 14C and 3H were discovered.
Before the second world war, the cyclotron was the main producer of
radionuclides since the neutron sources at that time were very weak. However,
with the development of the nuclear reactor, that situation changed. Suddenly,
a strong neutron source was available, which could easily produce almost
unlimited amounts of radioactive nuclides including biologically important
elements, such as 3H, 14C, 32P and 35S, and clinically interesting radionuclides,
such as 60Co (for external radiotherapy) and 131I, for nuclear medicine. After
the war, a new industry was born which could deliver a variety of radiolabelled
compounds for research and clinical use at a reasonable price.
However, accelerator produced nuclides have a special character, which
makes them differ from reactor produced nuclides. Today, their popularity is
increasing again. Generally, reactor produced radionuclides are most suitable for
laboratory work, whereas accelerator produced radionuclides are more useful
clinically. Some of the most used radionuclides in nuclear medicine, such as
In, 123I and 201Tl, and the short lived radionuclides, 11C, 13N, 15O and 18F, used
for positron emission tomography (PET), are all cyclotron produced.
4.1.2. Nuclide chart and line of nuclear stability
During the late 19th century, chemists learned to organize chemical
knowledge into the periodic system. Radioactivity, when it was discovered,
conflicted with that system. Suddenly, various samples, apparently with the
same chemical behaviour, were found to have different physical qualities such
as half-life, emitted type of radiation and energy. The concept of isotopes or
elements occupying the same place in the periodic system (from the Greek
(isos topos) meaning same place) was introduced by Soddy
1913, but a complete explanation had to await the discovery of the neutron by
Chadwick in 1932.



The periodic system was organized according to the number of protons

(atom number) in the nucleus, which is equal to the number of electrons to
balance the atomic charge. The nuclide chart consists of a plot with the number
of neutrons in the nucleus on the x axis and the number of protons on the y axis

Number of protons


Number of neutrons
FIG.4.1. Chart of nuclides. The black dots represent 279 naturally existing combinations of
protons and neutrons (stable or almost stable nuclides). There are about 2300 proton/neutron
combinations that are unstable around this stable line.

Figure4.2 shows a limited part of the nuclide chart. The formal notation
for an isotope is ZA X , where X is the element name (e.g. C for carbon), A is the
mass number (A=Z + N), Z is the number of protons in the nucleus (atom
number) and N the number of neutrons in the nucleus.
The expression above is overdetermined. If the element name X is known,
so is the number of protons in the nucleus, Z. Therefore, the simplified notation
X is commonly used.
Some relations of the numbers of protons and neutrons have special names
such as:
Isotopes: the number of protons is constant (Z=constant).
Isotones: the number of neutrons is constant (N=constant).
Isobars: The mass number is constant (A=constant).


Of these expressions, only the isotope concept is generally used. It is

important to understand that whenever the expression isotope is used, it must
always be related to a specific element or group of elements, for example,
isotopes of carbon (e.g. 11C, 12C, 13C and 14C).

Number of protons


















































































Number of neutrons
FIG.4.2. A part of the nuclide chart where the lightest elements are shown. The darkened
fields represent stable nuclei. Nuclides to the left of the stable ones are radionuclides deficient
in neutrons and those to the right, rich in neutrons.

In the nuclide chart (Fig.4.1), the stable nuclides fall along a monotonically
increasing line called the stability line. The stability of the nucleus is determined
by competing forces: the strong force that binds the nucleons (protons and
neutrons) together and the Coulomb force that repulses particles of like charge,
e.g. protons. The interplay between the forces is illustrated in Fig.4.3.
For best stability, the nucleus has an equal number of protons and
neutrons. This is a quantum mechanic feature of bound particles and in Fig.4.1
this is illustrated by a straight line. It is also seen that the stability line follows
the straight line for the light elements but that there is considerable deviation
(neutron excess) for the heavier elements. The explanation is the large Coulomb
force in the heavy elements which have many protons in close proximity. By
diluting the charge by non-charged neutrons, the distance between the charges
increases and the Coulomb force decreases.



FIG.4.3. Between the proton and a neutron, there is a nuclear force that amounts to
2.225MeV. The nucleons form a stable combination called deuterium, an isotope of hydrogen.
In a system of two protons, the nuclear force is equally strong to a neutronproton, but the
repulsive Coulomb forces are stronger. Thus, this system cannot exist. The nuclear force
between two neutrons is equally strong and there is no Coulomb force. Nevertheless, this
system cannot exist due to other repulsive forces, a consequence of the rules of pairing quarks.

4.1.3. Binding energy, Q-value, reaction threshold

and nuclear reaction formalism
There are no barriers and no repulsive forces between a free proton and
neutron, and they can fuse at low kinetic energies to form a deuterium nucleus,
which has a weight somewhat smaller than the sum of the free neutron and
proton weights. This mass difference can be converted into energy using
Albert Einsteins formula E=mc2 and is found to be 2.2MeV. This is also the
energy released as a photon in the reaction. To separate the two nucleons in
the deuterium nucleus, at least 2.2MeV have to be added. The energy gained
or lost in a nuclear reaction is called the Q-value. In a somewhat more complex
reaction, 14N(p, )11C, the Q-value is calculated as the difference between the
summation of the mass of the particles before the reaction (p, 14N) from the mass
of the particles after the reaction (, 11C). It should be noted that it is the mass
of the nucleus and not the atomic mass that is used. Using a Q-value calculator1,
the Q-value for the reaction 14N(p, )11C is 2923.056keV. This means that the
proton, when it reaches the 14N nucleus, has to have a kinetic energy of at least
2.93MeV in order to make the reaction possible.
However, before it hits the nucleus, the proton has to overcome the barrier
created by the repulsive Coulomb force between the proton and the positive 14N
nucleus. During the passage, the proton loses some energy and the starting value,
called the threshold value, must then exceed the Q-value. The same calculator
gives the threshold value of 3.14MeV for the 11C production reaction.
The reaction energy (the Q-value) is positive for exothermal reactions
(spontaneous reactions) and negative for endothermal reactions. Since all radioactive
decays are spontaneous, they need to have positive Q-values. Some reactions used

For example, http://nucleardata.nuclear.lu.se/database/masses/



to produce radionuclides, mainly those that are based upon thermal neutrons, have
positive Q-values but reactions based on positive particles usually have negative
Q-values, e.g. extra energy needs to be added to get the reaction going.
4.1.4. Types of nuclear reaction, reaction channels and cross-section
As seen in Fig.4.1, the radionuclides to the right of the stability line have
an excess of neutrons compared to the stable elements and they are preferentially
produced by irradiating a stable nuclide with neutrons. The radionuclides to
the left are neutron deficient or have an excess of charge and, hence, they are
mainly produced by irradiating stable elements by a charged particle, e.g. p or d.
Although these are the main principles, there are exceptions.
Usually, the irradiating particles have a large kinetic energy that is
transferred to the target nucleus to enable a nuclear reaction (the exception
being thermal neutrons that can start a reaction by thermal diffusion). Figure4.4
shows schematically an incoming beam incident upon the target, where it may
be scattered and absorbed. It can transfer its energy totally or partly to the target
nucleus and can interact with parts of or the whole of the target nucleus. Since the
produced activity should be high, the target is also usually thick.

I0Nd d/d

Incident beam


Scattered particles

FIG.4.4. Target irradiation. A nuclear physicist is usually interested in the particles coming
out, their energy and angular distribution, but the radiochemist is mainly interested in the
transformed nuclides in the target.

In radionuclide production, the nuclear reaction always involves a change

in the number of protons or neutrons. Reactions that result in a change in the


number of protons are preferable because the product becomes a different

element, facilitating chemical separation from the target, compared to an (n, )
reaction, where the product and target are the same.
Neutrons can penetrate the target at down to thermal energies. Charged
particles need to overcome the Coulomb barrier to penetrate the nucleus (Fig.4.5).

FIG.4.5. General cross-sectional behaviour for nuclear reactions as a function of the incident
particle energy. Since the proton has to overcome the Coulomb barrier, there is a threshold that
is not present for the neutron. Even very low energy neutrons can penetrate into the nucleus to
cause a nuclear reaction.

The parameter cross-section is the probability of a certain nuclear reaction

happening and is expressed as a surface. It is the probability that a particle will
interact per unit surface area of target. The geometrical cross-section of a uranium
nucleus is roughly 1028 m2, and this area has also been taken to define the unit
for cross-section barn (b). This is not an International System of Units unit but is
commonly used to describe reaction probabilities in atomic and nuclear physics.
For fast particle reactions, the probability is usually less than the
geometrical cross-section area of the nucleus, with probabilities in the range
of millibarns. However, the probability of a hit is a combination of the area of
both the nucleus and the incoming particle. The Heisenberg uncertainty principle
states that the position and the momentum of particles cannot be simultaneously
known to arbitrarily high precision. This implies that particles of well defined
but low energy, such as thermal neutrons, will have a large uncertainty in their
position. One may also say that they are increasing in size and nuclear reactions
involving thermal neutrons may have very large cross-sections, sometimes of the
order of several thousand barns.


The general equation for a nuclear reaction is:

where a is the incoming particle and A is the target nucleus in the ground state (the
entrance channel). Depending on the energy and the particle involved, several
nuclear reactions may happen, each with its own probability (cross-section). Each
nuclear reaction creates an outgoing channel, where b is the outgoing particle
or particles and B is the rest nucleus. Q is the reaction energy and can be both
negative and positive.
A common notation of a nuclear reaction is A(a, b)B. If the incoming
particle is absorbed, there is a capture process type (n, ) and in a reaction of type
(p, n) charge exchange occurs. If many particles are expelled, the reaction can
be referred to as (p, 3n). Each such reaction is called a reaction channel and is
characterized by an energy threshold (an energy that makes the nuclear reaction
possible, opens up the channel) and a probability (cross-section) varying with the
incoming particle energy. A schematic illustration of different reaction channels
opened in proton irradiation is given in Fig.4.6.

FIG.4.6. A schematic figure showing some reaction channels upon proton irradiation.

Different reaction mechanisms can operate in the same reaction channel.

Here, two ways are differentiated:
The formation of a compound nucleus;
Direct reactions.


The compound nucleus has a large probability to be formed in a central

hit of the nucleus and is preferable at low energies close to the energy threshold
of the reaction channel. Here, the incoming particle is absorbed and an excited
compound nucleus formed. This compound nucleus will rapidly (~1019 s)
undergo decay (fragment) with the isotropic emission of neutrons and rays.
Direct reactions preferentially occur at the edge of the nucleus or at high energies.
The incoming energy is directly transferred to a nucleon (knock-on reaction)
giving two outgoing particles. The outgoing particles usually have high energy
and are emitted in about the same direction as the incoming particle.

FIG.4.7. A schematic view of particle energy variations of a cross-section for direct nuclear
reactions and for forming a compound nucleus.

The production of radionuclides is due to a mixture of these two reaction

types. Their probability varies with energy in different ways. The direct
reactions are heavily associated with the geometrical size of the nucleus, and
the cross-section is usually small and fairly constant with energy. The highest
probability of forming a compound nucleus is just above the reaction threshold as
seen in Fig.4.7.
There are two major ways to produce radionuclides: using reactors
(neutrons) or particle accelerators (protons, deuterons, particles or heavy ions).
Since the target is a stable nuclide, either a neutron-rich radionuclide (reactor


produced) or a neutron deficient radionuclide (accelerator produced) is generally

4.2.1. Principle of operation and neutron spectrum
A nuclear reactor is a facility in which a fissile atomic nucleus such as 235U,
Pu or 241P absorbs a low energy neutron and undergoes nuclear fission. In the
process, fast neutrons are produced with energies from about 10MeV and below
(the fission neutron spectrum). The neutrons are slowed down in a moderator
(usually water) and the slowed down neutrons start new fissions. By regulating
this nuclear chain reaction, there will be a steady state of neutron production with
a typical neutron flux of the order of 1014 neutrons cm2 s1.
Since neutrons have no charge and are, thus, unaffected by the Coulomb
barrier, even thermal neutrons (0.025 eV) can enter the target nucleus and
cause a nuclear reaction. However, some nuclear reactions, depending upon the
cross-section, require fast neutrons (energy < 10MeV).
A reactor produces a neutron cloud in which the target is placed so that it
will be isotropically irradiated. Placing the target in different positions exposes
it to neutrons of different energy. Usually, the reactor facility has a pneumatic
system for placing targets at predefined positions. One has to consider the heat
that is generated in the reactor core, since the temperature at some irradiation
positions may easily reach 200C. The reactor is characterized by the energy
spectrum, the flux (neutrons cm2 s1) and the temperature at the irradiation
Most reactors in the world are for energy production and, for safety reasons,
cannot be used for radionuclide production. Usually, only national research
reactors are flexible enough for use in radioisotope production.

4.2.2. Thermal and fast neutron reactions

The most typical neutron reaction is the (n, ) reaction in which a thermal
neutron is captured by the target nucleus forming a compound nucleus. The
decay energy is emitted as a prompt ray. A typical example is the reaction
Co(n, )60Co that produces an important radionuclide used in external therapy.
However, since the produced radionuclide is of the same element as the target, the
specific activity a, i.e. the radioactivity per mass of the sample, is low. This type
of nuclear reaction is of little interest when labelling radiopharmaceuticals. In
light elements, other nuclear reactions resulting from thermal neutron irradiation
are possible, such as (n, p). Table4.1 lists possible production reactions for some
biologically important radionuclides.




Type of neutrons

Nuclear reaction




Co(n, ) Co


N(n, p)14C



S(n, p) P


Cl(n, )32P


Half-life T1/2

Cross-section (mb)

5.3 a


5730 a


25 d


24 d



15 h


Cl(n, )32P

14 d



Mg(n, p) Na


Nuclear reactions with thermal neutrons are attractive for many reasons.
The yields are high due to large cross-sections and the high thermal neutron fluxes
available in the reactor. In some cases, the yields are sufficiently high to use these
reactions as the source of charged secondary particles, e.g. 6Li(n, )3H for the
production of high energy 3H ions, which can then be used for the production of
F by 16O(3H, n)18F. The target used is 6LiOH, in which the produced 3H ions
will be in close contact with the target 16O. A drawback of this production is that
when the target is dissolved the solution is heavily contaminated with 3H water
that might be difficult to remove. Today, with an increasing number of hospital
based accelerators, there is little need of neutron produced 18F.
Another reactor produced neutron deficient radionuclide is 125I:

Xe(n, )125Xe (T1/2=17 h) 125I (T1/2=60 d)

This is currently the common way of producing high quality 125I. A

drawback is that 124Xe has a natural abundance of 0.1%. To increase the
production yield, one needs to work with expensive enriched targets. However,
these can be reused several times. This is an example of a generator system where
the mother is shorter lived than the daughter. Although there is no need to make a
separation between the mother and daughter, the target, after irradiation, has to be
stored for some days to allow the decay of 125Xe to be complete. The expensive
target gas 124Xe is carefully removed and the 125I is washed out from the walls of
the target capsule.
Many reactor produced radionuclides emit high energy particles that
contribute to the absorbed dose (but not the imaging signal) to patients, which is
a drawback in diagnostic procedures. However, a few emitting isotopes result
in daughter nuclei that emit rays with long de-excitation times (metastable


excited levels), instead of the more common prompt (1014 s) emission. Such
radioisotopes are suitable for nuclear medicine imaging, since they principally
yield radiation, with some electron emission, a consequence of internal
conversion. The most commonly used radionuclide in nuclear medicine, 99mTc,
is of this type. The m after the atomic mass signifies that this is the metastable
version of the radionuclide.
In radionuclide therapy, in contrast to diagnostic applications, the emission
of high energy radiation is desirable. Most radionuclides for radiotherapy are,
therefore, reactor produced. Examples include 90Y, 131I and 177Lu. A case of
interest to study is 177Lu, which can be produced in two different ways using
thermal neutrons. The most common production route is still the (n, ) reaction
on 176Lu, which opposes two conventional wisdoms in practical radionuclide
production for biomolecular labelling:
(a) Not to use a production that yields the same product element as the target
since it will negatively affect the labelling ability due to the low specific
(b) Not to use a target that is radioactive.
However, 176Lu is a natural radioactive isotope of lutetium with an
abundance of 2.59%. Figure4.8 shows how 177Lu needs to be separated from
the dominant 175Lu to decrease the mass of the final product. This method of
production works because the high cross-section (2020 b) of 176Lu results in a
high fraction of the target atoms being converted to 177Lu, yielding an acceptable
specific radioactivity of the final product.
On the right of Fig.4.8, an indirect way to produce 177Lu from 176Yb is also
shown. This method of production utilizes a generator nuclide 177Yb, produced
by an (n, ) reaction, which then decays to 177Lu. In principle, by chemically
separating lutetium from ytterbium, one would obtain the highest possible specific
radioactivity. However, the chemical separation between two lanthanides is not
trivial and, thus, it is difficult to obtain 177Lu without substantial contamination of
the target material Yb that may compete in the labelling procedure. Furthermore,
the cross-section for this reaction is almost a thousandfold lower, resulting in a
much lower product yield.
Reactions involving fast neutrons usually have cross-sections that are of the
order of millibarns, which, coupled with the much lower neutron flux at higher
energy relative to thermal neutron fluxes, leads to lower yields. However, there
are some important radionuclides, e.g. 32P that have to be produced this way.
Figure4.9 gives the details of this production.



Abundance (%)


3.78 1010 a

6.734 d

28.4 min


Abundance (%)

3.78 1010 a
4.185 d

6.734 d

28.4 min


1.911 h

FIG. 4.8. Production of 177Lu from 176Lu (left) and from 176Yb (right).

Abundance (%)

Abundance (%)

32P S
14.26 d

25.4 d

Abundance (%)
Abundance (%)


14.26 d


25.4 d

Neutron energy

FIG. 4.9. Data for the production of 32P in the nuclear reaction 32S(n, p)32P. The reaction
threshold is 0.51 MeV. From the cross-section data, it can be seen that there is no substantial
yield until an energy of about 2 MeV. The yield is an integration of the cross-section data and
the neutron energy spectrum. A practical cross-section can be calculated to about 60 mb.

4.2.3. Nuclear fission, fission products

Uranium-235 is not only used as fuel in a nuclear reactor but it can also be
used as a target to produce radionuclides. Uranium-235 irradiated with thermal
neutrons undergoes fission with a cross-section of 586 b. The fission process
results in the production of two fragments of 235U nucleus plus a number of free
neutrons. The sum of the fragments mass will be close to the mass of 235U, but
they will vary according to Fig. 4.10.
The masses of the 99Mo and 134Sn produced in the reaction:

U + n 236U 99Mo + 134Sn + 3n

are marked in Fig. 4.10. Some medically important radionuclides are produced
by fission, such as 90Y (therapy) and 99mTc (diagnostic). They are not produced
directly but by a generator system:

Sr (28.5 a) 90Y (2.3 d) and 99Mo (2.7 d) 99mTc (6 h)



U-235, fission yield

Yield/mass number (%)

Mass 134
Mass 99

Mass number
FIG.4.10. The yield of fission fragments as a function of mass.

The primary radionuclides produced are then 90Sr and 99Mo, or more
precisely the mass numbers 90 and 99.
Another important fission produced radionuclide in nuclear medicine,
both for diagnostics and therapy, is 131I. The practical fission cross-section for
this production is the fission cross-section of 235U multiplied by the fraction
of fragments having a mass of 131 or 5860.029=17 b. The probability of
producing a mass of 131 is 2.9% per fission. Iodine-131 is the only radionuclide
with a mass of 131 that has a half-life of more than 1 h, meaning that all of the
others will soon have decayed to 131I.
Charged particles, unlike neutrons, are unable to diffuse into the nucleus,
but need to have sufficient kinetic energy to overcome the Coulomb barrier.
However, charged particles are readily accelerated to kinetic energies that open
up more reaction channels than fast neutrons in a reactor. An example is seen in
Fig.4.11 that also illustrates alternative opportunities with: p, d, 3He and 4He or
, to produce practical and economic nuclear reactions.




i(p, 5n) 123 Xe 123 i


Xe(p, np) 123 Xe 123 i


Te(p, n) 123 i


Te(d, n) 123 i


Te(p, 2n) 123 i


sb( 4 he, 2n) 123 i


sb( 3 he, n) 123 i


sb( 3 he, 3n) 123 i

FIG.4.11. Various nuclear reactions that produce 123I. All of the reactions have been tried
and can be performed at relatively low particle energies. The 123Xe produced in the first two
reactions decays to 123I with a half-life of about 2 h. In the first reaction, the 123Xe is separated
from the target and then decays, while in the second reaction, the 123I is washed out of the
target after decay.

An accelerator in particle physics can be huge, as in the European

Organization for Nuclear Research (CERN), with a diameter of more than 4 km.
Accelerators for radionuclide production are much smaller, as they need to
accelerate particles to much lower energies. The first reaction in Fig.4.11, where
five neutrons are expelled, is the most energy demanding, as it requires a proton
energy of about 510MeV=50MeV (the rule of thumb is that about 10MeV
are required per expelled particle). All of the other reactions require 20MeV or
less (seeTable4.2).
Another advantage with accelerator production is that it is usually easy to
find a nuclear reaction where the product is a different element from the target.
Since different elements can be separated chemically, the product can usually be
of high specific radioactivity, which is important when labelling biomolecules.
A technical difference between reactor and accelerator irradiation is
that in the reactor the particles come from all directions but in the accelerator
the particles have a particular direction. The number of charged particles is
often smaller and is usually measured as an electric current in microamperes
(1 A=61012 protons/s but 31012 alpha/s because of the two charges of the



Proton energy



Accelerated particles

Used for


Mainly single particle, p or d



Usually p and d




p and d, He and He may be available

PET, commercial production


Usually p only

Often placed in national

centres and have several

A drawback in accelerator production is that charged particles are stopped

more efficiently than neutrons; for example, 16MeV protons are stopped in
0.6mm Cu. A typical production beam current of 100 A hitting a typical target
area of 2cm2 will then put 1.6 kW in a volume of 0.1cm3, which will evaporate
most materials if not efficiently cooled. In addition, the acceleration of the beam
occurs in a vacuum but the target irradiation is at atmospheric pressure or in gas
targets at 1020 times over pressure. To separate the vacuum from the target, the
beam has to penetrate foils that will absorb some particle energy and they will
also become strongly activated.
4.3.1. Cyclotron, principle of operation, negativeandpositiveions
There are several types of accelerator, all of which can, in principle, be
used for radionuclide production. The dominant one for radionuclide production
is currently the cyclotron that was invented by Lawrence in the early 1930s.
Cyclotrons were first installed in hospitals in the 1960s, but during the past two
decades, hospital based small cyclotrons yielding 1020MeV protons have
become fairly common, especially with the rise of PET.
A cyclotron is composed of four systems:
(a) A resistive magnet that can create a magnetic field of 12 T;
(b) A vacuum system down to 105 Pa;
(c) A high frequency system (about 40 MHz) providing a voltage with a peak
value of about 40kV, although these figures can vary considerably for
different systems;
(d) An ion source that can ionize hydrogen to create free protons as well as
deuterium and particles.



FIG.4.12. The cyclotron principle. A negative ion is injected into the gap between D-shaped
magnets (Dees) (1). An alternating electric field is applied across the gap, which causes
The cyclotron
A negative
ion charge
is injected
to accelerate.
The magnetic
on a moving
it to the
shaped magnets
is applied
path of(1).
(2). The
is gap,
in causes the
to accelerate.
on aso moving
charge forces
it to bend into a
time the charged
the gap,
that it is continuously
of (3).
ever increasing radius (2). The applied electric field is reversed in
finally beingpath

direction each time the charged particle reaches the gap, so that it is continuously
accelerated, until finally being ejected (3).

The inside of a cyclotron is shown in Fig.4.12. The ion source is usually

the vacuum
and in
in larger
machines,it is usual to
In inside
currents but,
of several
then injected
from located
the outside
a central
internal The
for are
the chamber.
In accelerators
the magnet.
The main
of such
the ion
is to havefor
a slow
of gas
beam currents
as source
those dedicated
PET flow
facilities, it is
onto an The
modes of extraction
is made
into to
an beam
arc discharge.
species The
are extracted
depend aupon
or negative
ions are
of positive ions is
in a static
field. There
are several
the particles when
types of ion source with different operating characteristics. In modern accelerators,
in the outer orbits. Some beam current is invariably lost in the process and the deflector often
negative ions, protons or deuterium with two orbit electrons are usually used. These
becomes quite radioactive.
of the beam. accelerators usually accelerate negative ions that are more
the systems,
ion source
easily extracted. In these
a thinwith
is used that
will the
away the two orbit
in aa consequence,
magnetic field,
ions suddenly
move in change
a circular
electrons.is As
the the
to positive
since efficiency
and are chamber,
bent are
out two
of the
magnetic historically
field with ancalled
the shape of the letter D. These electrodes are hollow, which
be transported
further in
a beam
transport system or
ions tobeam
the electrodes.
is aoptical
gap between
will hit a production target directly. The target is usually separated from the vacuum by
metallic foils that are strong enough to withstand the pressure difference and the heat from the
beam energy, as it is transferred and absorbed by the foils. The reason why two
foils are used


the electrodes called the acceleration gap. If a voltage is applied between the
electrodes, the ions will experience the potential gradient when traversing the
gap between the electrodes. If the voltage polarity is switched at the correct rate,
the ions will be continuously accelerated when crossing the gap, thus resulting
in an increase in the ions energy and velocity. As their velocity increases, the
ions will move into a circular orbit of increasing radius. The time taken for the
ions to return to the gap is independent of their radius in accelerators <30MeV.
For the cyclotron to operate correctly, it is necessary for the frequency of the
electric field across the Dees to be the same as the frequency of the circulating
ions, so that the polarity changes upon each traversal of the ions across the Dees.
In commercial accelerators, with high beam currents of several
milliamperes, it is usual to have an internal target for radionuclide production
located inside the chamber. In accelerators with lower beam currents <100 A,
such as those dedicated for PET hospital facilities, it is more common to extract
the beam onto an external target system. The modes of extraction depend upon
whether positive or negative ions are accelerated. Extraction of positive ions is
made by using a deflector that applies a static electric field which acts upon the
particles when in the outer orbits. Some beam current is invariably lost in the
process and the deflector often becomes quite radioactive.
Modern proton/deuterium accelerators usually accelerate negative ions
that are more easily extracted. In these systems, a thin carbon foil is used
that will strip away the two orbit electrons. As a consequence, the particles
suddenly change from negative to positive charge and are effectively bent out
of the magnetic field with an almost 100% extraction efficiency and with little
The extracted beam can either be transported further in a beam optical
transport system or will hit a production target directly. The target is usually
separated from the vacuum by metallic foils that are strong enough to withstand
the pressure difference and the heat from the beam energy, as it is transferred
and absorbed by the foils. The reason why two foils are used is that the heat
produced by the beam passage has to be removed, which is facilitated by a flow
of helium gas between the foils. Helium is preferred as the cooling medium
since no induced activity will be produced in this gas.
4.3.2. Commercial production (low and high energy)
If the proton energy is >30MeV, the particles tend to be relativistic,
i.e. their mass and their cycle time in orbit increase. A constant frequency of the
accelerating electric field would cause the ions to come out of phase. This can
be compensated for either by increasing the magnetic field as a function of the
cyclotron radius (isochronic cyclotrons) or by decreasing the radiofrequency


during acceleration (synchrocyclotrons). Such accelerators tend to be more

complex and expensive and, for this reason, 30MeV is a typical energy for
commercial accelerators that need to have large beam currents and to be both
reliable and cost effective.
Commercial accelerators usually run beam currents of several milliamperes.
Since it is technically difficult to extract such high beam currents due to heating
problems in the separating foils, most commercial accelerators use internal
targets, i.e. targets that are placed inside the cyclotron vacuum as shown
schematically in Fig.4.13.

FIG.4.13. Schematic image of an internal target. The target material is usually thin (a few
tenths of a micrometre) and evaporated on a thicker backing plate. The target ensemble is
water cooled on the back. An advantage is that the beam is spread out over a large area which
facilitates cooling.

Many patients in nuclear medicine undergo single photon emission

computed tomography (SPECT) investigations. Besides reactor produced
Tc, commercial cyclotrons commonly produce 67Ga, 111In, 123I and 201Tl. In
addition, some PET radionuclides, such as 124I, are becoming commercially
available. Increasing demand for the 68Ge/68Ga generator has also led to
commercial production of the cyclotron produced mother nuclide 68Ge. Only a
few radionuclides of medical interest require production energies above 30MeV.
A limited number of high energy accelerators with high beam currents, usually
at national physics laboratories, have the capacity for the production of, for
example, 52Fe and 61Cu and other isotopes used for research activities.
4.3.3. In-house low energy production (PET)
Commercial accelerators dedicated to PET radioisotope production
are limited both in energy (<20 MeV) and in beam current (<100 A). Many
production routes utilize gases or water as target materials and, therefore, external


targets are to be preferred. Owing to the relatively low beam current, extraction is
not a problem. Since internal targets need to be taken in and out of the cyclotron
vacuum, they are not usually implemented in PET cyclotrons.
The importance of choosing the right reaction and target material is crucial
and is illustrated by the production of 18F. There are several nuclear reactions that
can be applied (Table4.3).
OF 18F

Ne(d, )18F

The nascent 16F will be highly reactive. In the noble gas Ne, it will diffuse
and stick to the target walls; difficult to extract


Same as above; in addition, the abundance of 21Ne is low (0.27%) and needs


The product and target are the same element; poor specific radioactivity


Cheap target but accelerators that can accelerate particles to 35MeV are
expensive and not common

Ne(p, )18F
F(p, d)18F
O(, d)18F


O(d, )18F



O(p, n) F

Small cross-section and no practical yields can be obtained

Expensive enriched target material but the proton energy is low (low cost
accelerator), which makes this the nuclear reaction of choice

Not only the nuclear reaction is important, but also the chemical
composition of the target. To irradiate 18O as a gas would be the purest target
(only target nuclide present) but handling a highly enriched gas in addition to
the hot-atom chemistry is complicated. Still, for some applications, this might
be the best choice. To irradiate 18O as an oxide and a solid target is possible
but the process following irradiation to dissolve the target and to chemically
separate 18F is complex, has a low yield and other elements in the oxide could
potentially contribute unwanted radioactivity. Enriched 18O water is a target of
choice as 18O is the dominant nucleus and hydrogen does not contribute to any
unwanted radioactivity. There is usually no need for target separation as water
containing 18F can often be directly used in the labelling chemistry. The target
water can also, after being diluted with saline, be injected directly into patients,
e.g. 18F-fluoride for PET bone scans. Water targets will produce 18F-fluoride for
use in stereospecific nucleophilic substitutions. An alternative production route is
neon gas production, 20Ne(d, )18F. Adding 19F2 gas to the neon as a carrier yields
18 19
F F that can be used for electrophilic substitution. Adding carrier lowers the
specific radioactivity of the labelled product.



A problem is the heat generated when the beam is stopped in a few

millilitres of target water. High pressure targets that force the water to remain in
the liquid phase can overcome some of these problems but production is usually
limited to beam currents <40 A. Gas and solid targets are advantageous as they
can withstand higher beam currents.
There are also several options for the production of 11C. These include:
B(d,n)11C, 11B(p,n)11C and 14N(p,)11C. The reactions on boron are made as
solid target irradiations while the reaction on nitrogen is a gas target application.
The routine production routes of common positron emitters associated with
PET are summarized in Table4.4.




Nuclear reaction


Yield (GBq)

N(d, n) O gas target


O(p, )13N liquid target



N(p, ) C gas target



O(p, n) F liquid target


Oxygen-15 is produced by deuteron bombardment of natural nitrogen

through the 14N(d,n)15O nuclear reaction. An alternative is the 15N(p, n)15O
reaction if a deuterium beam is not available. In this case, the target needs to
be enriched. In the nitrogen target, 15O-labelled molecular oxygen is produced
directly. Direct production of 11C-labelled carbon dioxide is possible by mixing
the target gas with 5% natural carbon dioxide as a carrier. Water labelled with 15O
is preferably made by processing 15O-labelled molecular oxygen.
Carbon-11 is produced by proton bombardment of natural nitrogen. By
adding a small amount of oxygen to the target gas (<0.5%), carbon dioxide
(11CO2) will be produced. Adding 5% hydrogen to the target will produce
methane (11CH4).
Liquid targets are today by far the most popular and widely used for the
production of 13N. The reaction of protons on natural water produces nitrate
and nitrite ions, which can be converted to ammonia by reduction. Water targets
can also be used to form ammonia directly with the addition of a reducing
agent, e.g. ethanol or hydrogen.



4.3.4. Targetry, optimizing the production regarding

yield and impurities, yield calculations
When the nucleus is hit by an energetic particle, a complex interplay
between physical and statistical laws determines the result. Important parameters
are the entrance particle energy, the target thickness and the reaction channel
cross-sections for the particle energies in the target. Computer codes such as
ALICE and TALYS are available to calculate the size and the energy dependence
of the cross-section for a certain reaction channel but they are not easy to apply;
hence, caution should be exercised when interpreting the results from such codes.
However, a rough estimation of the irradiating particle energy can be obtained
using a well known rule of thumb in radionuclide production (illustrated in
Cross-section (b)

As(p, n)75Se


As(p, 3n)73Se


As(p, 4n)72Se


Proton energy Ep (MeV)

FIG.4.14. Excitation functions of 75As(p, xn)72,73,75Se reactions. The optimal energy for the
production of 73Se is to use a proton energy of 40MeV that is degraded to 30MeV in the target.

The maximum cross-sections are found at about 10, 30 and 40MeV for the
(p, n), (p, 3n) and (p, 4n) reactions, respectively. Thus, it takes about 10MeV to
expel a nucleon, i.e. a proton of 50MeV can cover radionuclide productions that
involve the emission of about five nucleons. At low energy, there is a disturbing
production of 75Se and if excessively high proton energy is used, another


unwanted radionuclide impurity is produced, namely 72Se. The latter impurity

can be avoided completely by restricting the proton energy to an energy lower
than the threshold for the (p, 4n) reaction. The impurity that results from the (p, n)
reaction cannot be avoided but can beminimized by using a target thickness that
avoids the lower proton energies (having the highest (p, n) cross-sections).
Figure4.14 highlights the fact that the chosen production parameters are a
compromise. A proton range of 4030MeV uses the (p, 3n) cross-section well.
Some 72Se contamination is acceptable in order to increase the yield of 73Se. An
important factor is the half-life of 75Se (T1/2=120 d), 73Se (T1/2=7.1 h) and 72Se
(T1/2=8.5 d). Sometimes, it is possible to wait for the decay of the radioactive
contaminants. Although not the case here, sometimes a long half-life contaminant
is not a serious disadvantage. If the product half-life is long, then there may
be little product decay over the target irradiation time compared to short lived
The practical set-up when undertaking radionuclide production is as follows.
A suitable As target is made and irradiated with 40MeV protons. The thickness
of the target is such that it decreases the proton energy to 30MeV. This then
gives a radioactivity yield of the desired radionuclide at the end of bombardment,
which is mainly dependent upon the beam current and the irradiation time. The
yield is usually expressed in gigabecquerels per microampere hours (GBq/Ah),
i.e. the produced radioactivity per time integrated beam current. If possible, it is
endeavoured to keep the radioactivity of the contaminants at low levels (<1%).
However, from the end of bombardment, the ratio of the product relative to any
long lived radio-contaminants begins to decrease.
Whenever a radionuclide (parent) decays to another radioactive nuclide
(daughter), this is called a radionuclide generator. Most natural radioactivity is
produced in generator systems starting with uranium isotopes and 232Th, and
involves about fifty radioactive daughters. Several radionuclides used in nuclear
medicine are produced by generator systems such as the 99Mo production of
Tc, which subsequently decays to 99Tc. The extremely long half-life of 99Tc
(T1/2=2.1105 a) means that 99mTc can be safely used as a clinical isotope
without any radiological concerns. In other nuclides, the creation of a radioactive
nuclide may be more important, e.g. the positron emitter 52Fe (T1/2=8 h) decays to
Mn (T1/2=21min) which is also a positron emitter. Furthermore, radionuclides
used in therapy may themselves be generators such as 211At (T1/2=7 h) decaying
to 211Po (T1/2=0.5 s) or 223Ra, which generates a series of relatively short lived
radioactive daughters in situ.


When talking about generators in nuclear medicine, a special case is usually

considered in which a long lived mother generates a short lived daughter, which
after labelling is administrated to the patient. Generally, this is a practical way
to deliver short lived radionuclides to hospitals which otherwise, for logistical
reasons, would not have been possible. The half-life should be sufficiently long
so that the radionuclide can be delivered to hospitals, and provide the radioactive
product for a number of patients over days or weeks. A typical example is the
Mo/99mTc generator (Fig.4.15), which produces the most used radionuclide in
nuclear medicine. The half-life of the parent (2.7 d) is adequate for transport and
delivery, and the daughter has a suitable half-life (6 h) for patient investigations.
The generator is used for about two to three half-lives of the parent (1 week) after
which time it is renewed.

FIG.4.15. Elution of a 99Mo/99mTc generator. The generator has a nominal activity of

1000MBq on day 0 (Monday). It is eluted daily, five times a week, yielding 1000, 780, 600,
470 and 360MBq.

4.4.1. Principles of generators

Generator systems require that the parent is a reactor or accelerator
produced by the methods described above and that the daughter radionuclide of
interest can easily be separated from the parent. The 99Mo/99mTc generator exhibits
these characteristics. Most commercial generators use column chromatography,
in which 99Mo is adsorbed onto alumina. Eluting the immobilized 99Mo on the
column with physiological saline elutes the soluble 99mTc in a few millilitres of



liquid. In fact, most generators in nuclear medicine use ion exchange columns in
much the same way due to its simplicity of handling.
In generator systems, the daughter radionuclide is formed at the rate at
which the parent decays, PNP. It also decays at the same rate, DND, as the parent,
once a state of transient equilibrium has been reached. The equations that describe the
relationship between parent and daughter are provided in Chapter1.
Another generator of increasing importance is 68Ge, which has a half-life
of 271 d that produces a short lived positron emitter 68Ga (T1/2=68min). This is
produced as a +3 ion that can be tagged, using a chelating agent such as DOTA,
to small peptides, e.g. 68Ga-DOTATOC. Owing to the long half-life of the mother,
the generator can be operated for up to two years and can be eluted every 5 h. One
problem with such a long lived generator is keeping it sterile, and furthermore,
the ion exchange material is exposed to high radiation doses that may reduce the
elution efficiency and the quality of the product.
The 90Sr/90Y generator is used to produce the therapeutic radionuclide 90Y.
This generator is not distributed to hospitals but is operated in special laboratories
on account of radiation protection considerations associated with the long lived
parent. The daughter, 90Y, has a half-life of 2.3 d which is adequate for transport
of the eluted 90Y to distant hospitals.
Rb (4.5 h)/81mKr (13.5 s) for ventilation studies and 82Sr (25.5 d)/82Rb (75 s)
for cardiac PET studies are examples of other generators with special requirements
due to the extremely short half-life of the eluted product. Recently,
generator systems producing emitters for therapy have become available,
e.g. 225Ac (10 d)/213Bi (45.6min).
During target irradiation, a few atoms of the wanted radionuclide are
produced within the bulk target material. The energy released in a nuclear
reaction is large relative to the electron binding energies and the radionuclide is,
therefore, usually born almost naked with no or few orbit electrons. This hot
atom will undergo chemical reactions depending on the target composition. In
a gas or liquid target, these hot atom reactions may even cause the activity to
be lost in covalent bonds to the target holder material. During irradiation, the
target is also heated and its structure and composition may change. A pressed
powder target may be sintered and become more ceramic, which makes it more
difficult to dissolve. The target may melt and the radioactivity may diffuse in the
target and even possibly evaporate. In designing a separation method, all of these
factors have to be considered. Fast, efficient and safe methods are required to



separate the few picograms of radioactive product from the bulk target material
which is present in gram quantities.
Separation of the radionuclide already starts in the target as demonstrated
in the production of 11CO2. Carbon-11 is produced in a (p, ) reaction on
nitrogen gas. To enable the production of CO2, some trace amounts of oxygen
gas (0.10.5%) are added. However, at low beam currents, mainly CO will be
formed, since the target will not be heated. At high beam currents, the CO will
be oxidized to the chemical form CO2. The separation, made by letting the target
through a liquid nitrogen trap, is simple and efficient. By adding hydrogen gas
instead, the product will be CH4.
The skill in hot-atom chemistry is to obtain a suitable chemical form of
the radioactive product, especially when working with gas and liquid targets.
Solid targets are usually dissolved and chemically processed to obtain the wanted
chemical form for separation.
4.5.1. Carrier-free, carrier-added systems
The concept of specific activity a, i.e. the activity per mass of a preparation,
is essential in radiopharmacy. If 100% of the product contains radioactive atoms,
often called the theoretical a, then the relationship between the activity A in
becquerels and the number of radioactive atoms N is given by N=A/, where is
the decay constant (1/s). The decay constant can be calculated from the half-life
T1/2 in seconds as =ln(2)/T1/2.
The specific activity a expressed as activity per number of radioactive atoms
is then A/N==ln(2)/T1/2. For a short lived radionuclide, a will be relatively
large compared to a long lived isotope. For example, a for 11C (T1/2=20min) is
1.5108 times larger than for 14C (T1/2=5730 a).
The specific activity a expressed in this way is a theoretical value that is
rarely obtained in practical work. When producing 11C, the target gas and target
holder will contain stable carbon that will dilute the radioactive carbon as well
as compete in the labelling process afterwards. A more empirical way to define
a is to divide the activity by the total mass of the element under consideration.
This value for 11C will usually be a few thousand times lower than the theoretical
value, while the production of 14C can come closer to the theoretical a.
In the labelling process, a is usually expressed as the activity per number
of molecules (a sum of labelled and unlabelled molecules). Instead of using the
number of atoms or molecules, it is common to use the mole concept by dividing
N by Avogadros number (NA=6.0221023). A common unit for a is then
gigabecquerels per micromole.
If the radioactive atoms are produced and separated from the target without
any stable isotopes, the process is said to be carrier-free. If stable isotopes are


introduced as being a contaminant in the target or in the separation procedure,

the process is said to have no carrier added, i.e. no stable isotope is deliberately
added. Both of these processes usually give a high final a. However, it may be
necessary to use a target of the same element or it may be necessary to add extra
mass of the same element in order for the separation process to work. In this case,
carrier is added deliberately and a will usually be low.
It should be noted that a carrier does not necessarily need to be of the same
element. When labelling a radiopharmaceutical with a chelator and metal ions,
any ion fitting into the chelator will compete. An example is labelling a peptide
with 111In, when the activity will usually be delivered as InCl3 in a weak acid. By
sampling the activity with a stainless steel needle, Fe ions will be released and
will probably completely ruin the labelling process by outnumbering the 111In
4.5.2. Separation methods, solvent extraction,
ion exchange, thermal diffusion
After irradiation, the small amount of desired radioactivity (of the order of
nanomoles) usually needs to be separated from the bulk of the target in a suitable
form for the following labelling process and at high a. The separation time should
be related to the half-life of the radionuclide and should take at most one half-life.
Solid targets usually have to be dissolved, which is simple for salts such as NaI
but more complicated for, for example, Ni foils where boiling aqua regia may
have to be used. To speed up this process, the Ni foil can be replaced by a pressed
target of Ni powder that will increase the metal surface and will speed up the
dissolving process.
In general, two principles are used: liquid extraction and ion exchange. In
liquid extraction, usually two liquids that do not mix are used, e.g. water and an
organic solvent. The target element and the produced activity of another element
should have different relative solubility in the liquids. The two liquids and the
dissolved target are mixed by shaking, after which two phases are formed. The
phase with a high concentration of the wanted radioactive product is sampled
and is usually separated again one or more times to reduce the target mass in that
fraction. The relative solubility can be optimized by varying the pH or by adding
a complexing agent.
In the ion exchange mechanism, an ion in the liquid phase (usually an
aqueous phase) is transferred to a solid phase (organic or ceramic material). To
maintain the charge balance, a counter ion is released from the solid phase. This
ion may be a hydrogen ion. In the ion exchange mechanism, the distribution ratio
is often a function of the pH. Furthermore, complexing agents can be used to
modify the distribution ratio. The dissolved target is adjusted to obtain the right


pH and other separation conditions, and is then put on to a column containing the
ion exchange material. The optimal separation conditions would be that the small
mass of desired radioactivity but not the bulk target material sticks to the column.
The column can then be small, and after washing and change of pH, the desired
activity can be eluted in a small volume. Under other conditions, large amounts
of ion exchange material have to be used to prevent saturation of binding sites
and leakage of the target material. This also means that large liquid volumes have
to be used, implying poorer separation. The two techniques are often performed
together by using liquid extraction to reduce the target mass, after which ion
exchange is used to make the final separation.
Occasionally, thermal separation techniques may be applied, which have
the advantage that they do not destroy the target (important when expensive
enriched targets are used) and that they lend themselves to automation. As an
example of such dry methods, the thermal separation of 76Br (T1/2=16 h) is
described. The target is Cu276Se, a selenium compound that can withstand some
heat. The nuclear reaction used is 76Se(p, n)76Br.
The process is as follows:
(a) The target is placed in a tube and heated, under a stream of argon gas, to
evaporate the 76Br activity by dry distillation (Fig.4.16);
(b) A temperature gradient is applied to separate the deposition areas of 76Br and
traces of co-evaporated selenide in the tube by thermal chromatography;
(c) The 76Br activity deposited on the tube wall is dissolved in small amounts
of buffer or water.

FIG.4.16. A schematic description of the 76Br separation equipment: (1) furnace, (2) auxiliary
furnace, (3) irradiated target, (4) deposition area of selenium, (5) deposition area of 76Br,
(6) gas trap.



Separation yields of 6070% are achieved by this method, with a separation

time of about 1 h. Since dry distillation permits the extraction of radiobromine
without destroying the target, the Cu2Se targets are reusable. Considering the
rather expensive 76Se-enriched target material, this is a practical prerequisite for
this type of production. The chemical form of the 76Br activity after separation,
analysed by ion exchange high performance liquid chromatography and thin-layer
chromatography, was almost exclusively found to be bromide.
4.5.3. Radiation protection considerations and hot-box facilities
Besides the desired activity, the irradiated target usually contains a number
of other radionuclides of varying elements, half-lives and energies. The
presence of such contaminants needs to be taken into account when planning
radiopharmaceutical labelling. An example is the production of 35S using the
reaction 35Cl(n, p)35S. At first glance, NaCl would be a suitable target due to the
low atomic weight of sodium, a single isotope (23Na) and a salt that is easy to
dissolve. However, 23Na has a huge thermal neutron cross-section for producing
Na, which has a half-life of 15h and abundant energies up to 2.75MeV. This
target would be extremely hot, demanding lead protection more than 30cm thick.
If instead KCl were used, the emitted radiation energy would be substantially
lower and decay times shorter.
After irradiation, the target is usually stored before processing to allow
any short lived radionuclides to decay. Depending on the half-life, this cooling
period can be fromminutes to months, but should not exceed one half-life of the
desired radionuclide. The place used for this depends on the source activity and
the energy and abundance of the emissions. Separation of fairly pure and
emitters may require just some distance and some plastic shielding, and can be
performed in a standard fume hood, while targets with a high emission need
significant lead shielding.
Handling reactor or accelerator produced radioactivity of the order of
several hundred gigabecquerels requires adequate radiation protection, usually
in the form of lead shields, hot-boxes, lead shielded fume hoods and laminar air
flow benches. Typical lead thicknesses required by common radionuclides are
indicated in Table4.5.
The radioactive target and the radionuclide separation is often the first
step in labelling a radiopharmaceutical. The hot-box then has to fulfil the
requirements both to protect the operator from the radiation and to protect the
pharmaceutical from the surroundings. The first step usually requires a negative
pressure hood to prevent eventual airborne radioactivity to leak out into the
laboratory, while the second step requires a high positive pressure to be applied
across the pharmaceutical to avoid contact with less pure air from the laboratory.



Dose rate (mSv/h) at 1 m per TBq










Thickness of lead shield (cm) giving 1 Sv/h























Calculations made with RadProCalculator (http://www.radprocalculator.com/).

These contradictory conditions are usually handled by having a box in the box,
i.e. the pharmaceutical is processed in a closed facility at over pressure placed in
the hot-box having low pressure. The classical hot-box design, with manipulators
to manually process the radioactivity remotely, as seen in Fig.4.17, is gradually
being replaced by lead protected chambers housing an automatic chemistry
system or a chemical robot making the pharmaceutical computer controlled.

FIG. 4.17. Examples of modern hot-box designs (courtesy of Von Gahlen Nederland B.V.).


FIG. 4.17. Examples of modern hot-box designs.

Department of Medical Physics,
University of the Free State,
Bloemfontein, South Africa
Measurement errors are of three general types: (i) blunders, (ii) systematic
errors or accuracy of measurements, and (iii) random errors or precision of
Blunders produce grossly inaccurate results and experienced observers
easily detect their occurrence. Examples in radiation counting or measurements
include the incorrect setting of the energy window, counting heavily contaminated
samples, using contaminated detectors for imaging or counting, obtaining
measurements of high activities, resulting in count rates that lead to excessive
dead time effects, and selecting the wrong patient orientation during imaging.
Although some blunders can be detected as outliers or by duplicate samples and
measurements, blunders should be avoided by careful, meticulous and dedicated
work. This is especially important where results will determine the diagnosis or
treatment of patients.
Systematic errors produce results that differ consistently from the correct
results by some fixed amount. The same result may be obtained in repeated
measurements, but overestimating or underestimating the true value. Systematic
errors are said to influence the accuracy of measurements. Measurement results
having systematic errors will be inaccurate or biased. Examples of a systematic
error are:
When an incorrectly calibrated ionization chamber is used for measurement
of radiation dose.
When during thyroid uptake studies with 123I the count rate of the reference
standard results in dead time losses. The percentage of thyroid uptake will
be overestimated.
When in sample counting the geometry of samples and the position within
the detector are not the same as in the reference sample.



When during blood volume measurements the tracer leaks out of the blood
compartment. The theory of the method assumes that the tracer will stay
in the blood compartment. The leaking of the tracer will consistently
overestimate the measured blood volume.
When in calculation of the ventricular ejection fraction during gated
blood pool studies the selected background counts underestimate the true
ventricular background counts, the ejection fraction will be consistently
Measurement results affected by systematic errors are not always easy to
detect, since the measurements may not be too different from the expected results.
Systematic errors can be detected by using reference standards. For example,
radionuclide standards calibrated at a reference laboratory should be used to
calibrate source calibrators to determine correction factors for each radionuclide
used for patient treatment and diagnosis.
Measurement results affected by systematic errors can differ from the true
value by a constant value and/or by a fraction. Using golden standard reference
values, a regression curve can be calculated. The regression curve can be used to
convert systematic errors to a more accurate value. For example, if the ejection
fraction is determined by a radionuclide gated study, it can be correlated with the
golden standard values.
Random errors are variations in results from one measurement to the next,
arising from actual random variation of the measured quantity itself, as well as
physical limitations of the measurement system.
Random error affects the reproducibility, precision or uncertainty in the
measurement. Random errors are always present when radiation measurements
are performed because the measured quantity, namely the radionuclide decay,
is a random varying quantity. The random error during radiation measurements
introduced by the measured quantity, that is the radionuclide decay, is
demonstrated in Fig.5.1. Figure5.1 shows the energy spectrum of a 57Co source
in a scattering medium and measured with a scintillation detector probe. The
energy spectrum represented by square markers is the measured energy spectrum
with random noise due to radionuclide decay. The solid line spectrum represents
the energy spectrum without random noise. The variation around the solid line of
the data points, represented by markers, is a result of random error introduced by
radionuclide decay.
The influence of the random error of the measurement system introduced
by the scintillation detector is also demonstrated in Fig.5.1. Cobalt-57 emits
photons of 122keV and with a perfect detection system all of the counts are
expected at 122keV. The measurements are, however, spread around 122keV
as a result of the random error introduced by the scintillation detector during the


detection of each photon. When a photon is detected with the scintillation

detector, the number of charge carriers generated will vary randomly. The varying
number of charge carriers will cause varying pulse heights at the output of the
detector and this variation determines the spread around the true photon energy
of 122keV. The width of the photopeak determines the energy resolution of the
detection system.

FIG.5.1. Energy spectrum of a

scintillation detector.


Co source in a scattering medium obtained with a

Random errors also play a significant role in radionuclide imaging. Here,

the random error as a result of the measured quantity, namely radionuclide decay,
will significantly influence the visual quality of the image. This is because the
number of counts acquired in each pixel is subject to random error. It is shown that
the relative random error decreases as the number of counts per pixel increases.
The visual effect of the random error as a result of the measured quantity is
demonstrated in Fig.5.2. Technetium-99m planar bone scans (acquired on a
256256 matrix) were acquired with a scintillation camera. Image acquisition
was terminated at a total count of 21, 87 and 748 kcounts. When the total number
of counts per image are increased, the counts per pixel increase and the random
error decreases, resulting in improved visual image quality. As the accumulated
counts are increased, the ability to visualize anatomical structures and, more
importantly, tumour volumes, significantly increases. The random error
introduced by the measuring system or imaging device, such as a scintillation
camera, also influences image quality. This is as a result of the energy resolution
and intrinsic spatial resolution of imaging devices that are influenced by random
errors during the detection of each photon. The energy resolution of the system


will determine the ability of the system to reject lower energy scattered photons
and improve image contrast.

FIG.5.2. The influence of random error as a result of radionuclide decay or counting statistics
is demonstrated for imaging. Technetium-99m posterior planar bone images (256256) using
a scintillation camera were acquired to total counts of 21, 87 and 748 kcounts.

It is possible for a measurement to be precise (small random error) but

inaccurate (large systematic error), or vice versa. For example, for the calculation
of the ejection fraction during gated cardiac studies, the selection of the
background region of interest (ROI) will be exactly reproducible when a software
algorithm is used. However, if the algorithm is such that the selected ROI does
not reflect the true ventricular background, the measurement will be precise but
inaccurate. Conversely, individual radiation counts of a radioactive sample may
be imprecise because of the random error, but the average value of a number of
measurements will be accurate, representing the true counts acquired.
Random errors are always present and play a significant role in radiation
counting and imaging. It is, therefore, important to analyse the random errors to
determine the associated uncertainty. This is done using methods of statistical
analysis. The remainder of the chapter describes methods of analysis.
The analysis of radiation measurements and imaging forms a subgroup
of general statistical analysis. In this chapter, the focus is on statistical analysis
for radiation counting and imaging measurements, although some methods
described will be applicable to a wider class of experimental data as described in
Sections 5.2, 5.3 and 5.5.




5.2.1. Measures of central tendency and variability Dataset as a list
Two measurements of the central tendency of a set of measurements are
the mean (average) and median. It is assumed that there is a list of N independent
measurements of the same physical quantity:
x1, x2, x3, .xixN
It is supposed that the dataset is obtained from a long lived radioactive
sample counted repeatedly under the same conditions with a properly operating
counting system. As the disintegration rate of the radioactive sample undergoes
random variations from one moment to the next, the number of counts recorded
in successive measurements is not the same as the result of random errors in the
The experimental mean x e of the set of measurements is defined as:
xe =

x 1 + x 2 + ... + x N

i =1


The following procedure is followed to obtain the median. The list of

measurements must first be sorted by size. The median is the middlemost
measurement if the number of measurements is odd and is the average of the
two middlemost measurements if the number of measurements is even. For
example, to obtain the median of five measurements, 7, 13, 6, 10 and 14, they are
first sorted by size: 6, 7, 10, 13 and 14. The median is 10. The advantage of the
median over the mean is that the median is less affected by outliers. An outlier is
a blunder and is much greater or much less than the others.
The measures of variability, random error and precision of a list of
measurements are the variance, standard deviation and fractional standard
deviation, respectively.



The variance e2 is determined from a set of measurements by subtracting

the mean from each measurement, squaring the difference, summing the squares
and dividing by one less than the number of measurements:
e2 =

( x 1 x e ) 2 + ( x 2 x e ) 2 + ........ + ( x N x e ) 2
N 1

= N11

(x x )


i =1

where N is the total number of measurements and x e is the experimental mean.

The standard deviation e is the square root of the variance:
e = e 2 (5.4)

The fractional standard deviation eF (fractional error or coefficient of

variation) is the standard deviation divided by the mean:
eF =

x e (5.5)

The fractional standard deviation is an important measure to evaluate

variability in measurements of radioactivity. The inverse of the fractional
standard deviation 1/eF in imaging is referred to as the signal to noise ratio. Dataset as a relative frequency distribution function
It is often convenient to represent the dataset by a relative frequency
distribution function F(x). The value of F(x) is the relative frequency with which
the number appears in the collection of data in each bin. By definition:
F ( x) =

number of occurrences of the value x in each bin

number of measurements (N )


The distribution is normalized, that is:

F (x) = 1 (5.7)
x =0



As long as the specific sequence of numbers is not important, the complete

data distribution function represents all of the information in the original dataset
in list format.
Figure5.3 illustrates a demonstration of the application of the relative
frequency distribution. The scintillation counter measurements appear noisy due
to the random error as a result of the measured quantity (the radionuclide decay)
(Fig.5.3(a)). The measurements fluctuate randomly above and below the mean
of 90 counts. A histogram (red bars) of the relative frequency distribution of
the fluctuations in the measurements can be constructed by plotting the relative
frequency of the measured counts (Fig.5.3(b)). The x axis represents the range of
possible counts that were measured in each of the bins with a six count interval.
The y axis represents the relative frequencies with which the particular count
values occur. The most common value, that is 26% of the measurements, is near
the mean of 90 counts. The values of the other measurements are substantially
higher or lower than the mean. The measured frequency distribution histogram
agrees well with the expected calculated normal distribution (blue curve).

Measurement counts

(a) Measured counts

(b) Relative frequency distribution

Measured counts
Calculated frequency distribution

Measured counts

Measurement number


FIG.5.3. One thousand measurements were made with a scintillation counter. (a)The graph
shows the variations observed for the first 50 measurements. (b)The graph (red bars) shows
the histogram of the relative frequency distribution for the measurements as well as the
expected calculated frequency distribution.

The relative frequency distribution is a useful tool to provide a quick visual

summary of the distribution of measurement values and can be used to identify
outliers such as blunders or the correct functioning of equipment.



Three measurements of the central tendency for a frequency distribution

are the mean (average), median and mode:
The mode of a frequency distribution is defined as the most frequent value
or the value at the maximum probability of the frequency distribution.
The median of a frequency distribution is the value at which the integral of
the frequency distribution is 0.5; that is, half of the measurements will be
smaller and half will be larger than the median.
The experimental mean x e using the frequency distribution function
can be calculated. The experimental mean is obtained by calculating the first
moment of the frequency distribution function. The equation for calculating the
mean can also be derived from the equation for calculating the mean for data
in a list (Eq.(5.2)). The sum of measurements

is equal to the sum of the

i =1

measurements in each bin in the frequency distribution function. The sum of the
measurements for each bin is obtained by multiplying the value of the bin i and
the number of occurrences of the value xi.

xe =

i =1


[value of bin (i)] [number of occurences of the value x i ]

i =1

x F (x )



i =1

The experimental sample variance can be calculated using the frequency

distribution function:
e2 =

N 1

(x x ) F (x )


i =1

The standard deviation and the fractional standard deviation are given by
Eqs(5.4) and (5.5).
The frequency distribution provides information and insight on the
precision of the experimental sample mean and of a single measurement.
Figure5.3 demonstrates the distribution of counting measurements around the



true mean ( x t ) . The value of the true mean is not known but the experimental
sample mean ( x e ) can be used as an estimate of the true mean ( x t ) :
( x t ) ( x e ) (5.10)

In routine practice, it is often impractical to obtain multiple measurements

and one must be satisfied with only one measurement. This is especially the
case during radionuclide imaging and nuclear measurements on patients. The
frequency distribution of the measurements will determine the precision of a
single measurement as an estimate of the true value. The probability that a single
measurement will be close to the true mean depends on the relative width or
dispersion of the frequency distribution curve. This is expressed by the variance
2 (Eq.(5.9)) or standard deviation of the distribution. The standard deviation
is a number such that 68.3% of the measurement results fall within of the true
mean x t .
Given the result of a given measurement x, it can be said that there is a
68.3% chance that the measurement is within the range x. This is called the
68.3% confidence interval for the true mean x t . There is 68.3% confidence that
x t is in the range x. Other confidence intervals can be defined in terms of the
standard deviation . They are summarized in Table5.1. The 50% confidence
interval (0.675) is referred to as the probable error of the true mean x t .
TABLE5.1. Confidence Levels in radiation Measurements

Confidence level for true mean x t

x t 0.675


x t 1.000


x t 1.640


x t 2.000


x t 3.000



Under certain conditions, the distribution function that will describe
the results of many repetitions of a given measurement can be predicted. A
measurement is defined as counting the number of successes x resulting from a
given number of trials n. Each trial is assumed to be a binary process in that only



two results are possible: the trial is either a success or not. It is further assumed
that the probability of success p is constant for all trials.
To show how these conditions apply in real situations, Table5.2 gives
four separate examples. The third example gives the basis for counting nuclear
radiation events. In this case, a trial consists of observing a given radioactive
nucleus for a period of time t. The number of trials n is equivalent to the number
of nuclei in the sample under observation, and the measurement consists of
counting those nuclei that undergo decay. We identify the probability of success
as p. For radioactive decay:
p = (1 e t ) (5.11)

where is the decay constant of the radionuclide.

The fifth example demonstrates the uncertainty associated with the energy
determination during scintillation counting. The light photons generated in the
scintillator following interaction with an incoming ray will eject electrons at the
photocathode of the photomultiplier tube (PMT). Typically, one electron ejected
for every five light photons results in a probability of success of 1/5.
Table5.2. Examples of Binary Processes

Definition of success

Probability of success p

Tossing a coin



Rolling a die

A six


Observing a given radionuclide

for time t

The nucleus decays during


(1 e lt )

Observing a given ray over

a distance x in an attenuating

The ray interacts with the

medium during observation

(1 e x )

Observing light photons

generated in a scintillator

An electron is ejected from the



5.3.1. Conditions when binomial, Poisson and

normal distributions are applicable
Three statistical models are used: the binomial distribution, the Poisson
distribution and the Gaussian or normal distribution. Figure5.4 shows the



distribution for the three models. The distributions were generated by using a
Microsoft Office Excel spreadsheet. Binomial distribution
This is the most general model and is widely applicable to all constant
p processes (Fig.5.4). Binomial distribution is rarely used in nuclear decay
applications. One example in which the binomial distribution must be used is
when a radionuclide with a very short half-life is counted with a high counting






FIG.5.4. Probability distribution models for successful event probability p=0.4 and
p=0.0001 for x = 10 and x =100, respectively. Poisson distribution

The model is a direct mathematical simplification of the binomial
distribution under conditions that the event probability of success p is small
(Fig.5.4). For nuclear counting, this condition implies that the chosen observation
time is small compared to the half-life of the source, or that the detection


efficiency is low. The Poisson distribution is an important distribution. When

the success rate is low, the true experimental distribution is asymmetric and a
Poisson distribution must then be used since the normal distribution is always
symmetrical. This is demonstrated in Fig.5.4 for p=0.0001 and x =10. Gaussian or normal distribution
The third important distribution is the normal or Gaussian, which is
a further simplification if the mean number of successes x is relatively large
(>30). At this level of success, the experimental distribution will be symmetrical
and can be represented by the Gaussian distribution (Fig.5.4). The Gaussian
model is widely applicable to many applications in counting statistics.
It should be emphasized that the distribution of all of the above models
becomes identical for processes with a small individual success probability p
and with a large enough number of trials such that the expected mean number of
successes x is large. This is demonstrated in Fig.5.4 for p=0.0001 and x =100.
5.3.2. Binomial distribution
Binomial distribution is the most general of the statistical models
discussed. If n is the number of trials for which each trial has a success probability
p, then the predicted probability of counting exactly x successes is given by:
P( x) =

p x (1 p) nx (5.12)
(n x)! x !

P(x) is the predicted probability distribution function, and is defined only

for integer values of n and x. The value of n! (n factorial) is the product of integers
up to n, that is 123 n. The values of x! and (n x)! are similarly
The properties of the binomial distribution are as follows:
The distribution is normalized:

P( x) = 1


x =0

The mean value x of the distribution using Eq.(5.8) is given by:



x =0




If Eq.(5.12) is substituted for P(x), the mean value x of the distribution is

given by:
x = pn (5.15)

The sample variance for a set of experimental data has been defined by
Eq.(5.9). By analogy, the predicted variance 2 is given by:
2 =

(x x )

P( x) (5.16)

x =0

If Eq.(5.12) is substituted for P(x), the predicted variance 2 of the

distribution will be:
2 = np(1 p) (5.17)

If Eq.(5.15) is substituted for np:

2 = x (1 p) (5.18)

The standard deviation is the square root of the predicted variance 2:

= np(1 p) = x (1 p)


The fractional standard deviation F is given by:

F =

np(1 p)
(1 p)

F =

x (1 p)

(1 p)

Equation (5.19) predicts the amount of fluctuation inherent in a given

binomial distribution in terms of the basic parameters, namely the number of
trials n and the success probability p, where x = pn . Application example of binomial distribution
The operation of a scintillation detector (Section 6.4) is considered. It
consists of a scintillation crystal mounted on a PMT in a light tight construction.
Firstly, when a ray interacts with the crystal, it generates n light photons.


Secondly, the light photons then eject x electrons from the photomultiplier
photocathode. Thirdly, these electrons are then multiplied to form a pulse that
can be further processed. For each ray that interacts with the scintillator, the
number of light photons n, electrons ejected x and multiplication vary statistically
during the detection of the different rays. This variation determines the energy
resolution of the system.
In this example, the second stage is illustrated, that is the ejection of
electrons from the photocathode. The variation or the standard deviation and
fractional standard deviation for the number of electrons x that are ejected can
be calculated using the binomial distribution as is given by Eqs(5.19) and (5.20).
The typical values for a scintillation counter are as follows. It is assumed
that the 142keV rays emitted by 99mTc are being counted. It is further assumed
that it uses 100 eV to generate a light photon in the scintillation crystal when a
ray interacts with the crystal. Therefore, if all of the energy of a single 142keV
photon is absorbed, n=142 000/100=1420 light photons will be emitted. It is
assumed that these light photons fall on the photocathode of the PMT to generate
x electrons for each ray absorbed. It is further assumed that five light photons
are required to eject one electron.
For the binomial distribution, the probability of a light photon ejecting
an electron is p=1/5 and the number of trials n will be the number of light
photons generated for each ray. This will be 1420. Equation (5.15) can be used
to calculate the predicted mean number of electrons ejected for each ray:
x = pn =

1420 = 284 electrons

The standard deviation (Eq.(5.19)) and relative standard deviation

(Eq.(5.20)) can be calculated using the binomial distribution:
= x (1 p) = 284(1 1 / 5) = 15

F =

(1 p)
(1 1 / 5)
= 0.053 (5.21)

Therefore, the contribution to the overall standard deviation at the electron

ejection stage at the photocathode is 5.3%. The variation in the number of
electrons will influence the pulse height obtained for each ray. The variation
in the pulse height during the detection of rays will determine the width of the


photopeak (Fig.5.1) and the energy resolution of the system (Sections 5.7.1 and
5.3.3. Poisson distribution
Many binary processes can be characterized by a low probability of
success for each individual trial. This includes nuclear counting and imaging
applications in which large numbers of radionuclides make up the sample or
number of trials, but a relatively small fraction of these give rise to recorded
counts. Similarly, during imaging, many rays are emitted by the administered
imaging radionuclide, for every one that interacts with the tissue. In addition,
during nuclear counting, many rays strike the detector for every single recorded
Under these conditions, the approximation that the probability p is small
(p 1) will hold and some mathematical simplifications can be applied to the
binomial distribution. The binomial distribution reduces to the form:
P( x) =

( pn) x e pn


The relation pn = x holds for this distribution as well as for the binomial
P( x) =

( x ) x e x


Equation (5.23) is the form of the Poisson distribution.

For the calculation of binomial distribution, two parameters are required:
the number of trials n and the individual success probability p. It is noted from
Eq.(5.23) that only one parameter, the mean value x , is required. This is a very
useful simplification because using only the mean value of the distribution, all
other values of the Poisson distribution can be calculated. This is of great help
for processes in which the mean value can be measured or estimated, but for
which there is no information about either the individual probability or size of the
sample. This is the case in nuclear counting and imaging.
The properties of the Poisson distribution are as follows. The Poisson
distribution is a normalized frequency distribution (seeEqs(5.6) and (5.7)):

P(x) = 1 (5.24)
x =0



The mean value or first moment for the Poisson distribution is calculated
by inserting the Poisson distribution (Eq.(5.22)) into the equation to calculate the
mean for a frequency distribution (Eq.(5.8)):


xP(x) = pn
x =0


This is the same result as was obtained for the binomial distribution.
The predicted variance of the Poisson distribution differs from that of the
binomial distribution and can be derived from Eqs(5.9) and (5.22):
2 =

(x x )

P( x) = pn (5.26)

x =0

From the result of Eq.(5.26), the predicted variance is reduced to the

important general equation:
2 = x (5.27)

The predicted standard deviation is the square root of the predicted variance
= 2 = x (5.28)

The predicted standard deviation of any Poisson distribution is just the

square root of the mean value that characterizes the same distribution.
The predicted fractional standard deviation F (fractional error or coefficient
of variation) is the standard deviation divided by the mean (Eq.(5.5)):
F =

= (5.29)

The fractional standard deviation is the inverse of the square root of the
mean value of the distribution.
Equations (5.28) and (5.29) are important equations and frequently find
application in nuclear detection and imaging.



5.3.4. Normal distribution

The Poisson distribution holds as a mathematical simplification to the
binomial distribution within the limit p < 1. If, in addition, the mean value of the
distribution is large (>30), additional simplification can generally be carried out
which leads to a normal or Gaussian distribution:
( xx ) 2

2 x

P( x) =
2 x


The distribution function is only defined for integer values of x.

Figure5.4 for x = 100 and p=0.0001 demonstrates that for these values
the normal distribution is identical to the Poisson and binomial distributions. The
normal distribution is always symmetrical or bell-shaped (Fig.5.4). It shares
the following properties with the Poisson distribution:
It is normalized (seeSection and Eqs(5.6) and (5.7)):

P(x) = 1 (5.31)
x =0

The distribution is characterized by a single parameter x = pn .

The predicted variance of the normal distribution is given by the mean of x:
2 = x (5.32)

The predicted standard deviation is the square root of the predicted variance
= 2 = x (5.33)

The predicted fractional standard deviation F (fractional error or coefficient

of variation) is the standard deviation divided by the mean (Eq.(5.5)):



The fractional standard deviation is the inverse of the square root of the
mean value of the distribution.


CHAPTER 5 Continuous normal distribution: confidence intervals

In experiments where the sample size is small, there are only a few discrete
outcomes. As the sample size increase, so does the number of possible sample
outcomes. As the sample size approaches infinity, there is, in effect, a continuous
distribution of outcomes. In addition, some random variables, such as height and
weight, are essentially continuous and have continuous distributions. In these
situations, the probability of a single event is not small as was assumed for the
discrete Poisson and normal distributions, and the equation = x does not
apply. The continuous normal distribution is given by:
1 xx

e 2
P( x) =


The properties (Fig.5.5) of the continuous normal distribution are:


It is a continuous, symmetrical curve with both tails extending to infinity.

All three measures of central tendency, mean, median and mode, are
It is described by two parameters: the arithmetic mean x and the standard
deviation .
The mean x determines the location of the centre of the curve and the
standard deviation represents the spread around the mean.

Number of standard deviations

FIG.5.5. The continuous normal distribution indicating the probability levels at different
standard deviations (SDs) from the mean.


Relative response


Distance (mm)
FIG.5.6. Line source response curve obtained from a scintillation camera fitted to a normal
distribution model. Image resolution is measured as the distance of the full width at half
maximum (FWHM) of the percentage response. The standard deviation (SD) is the half width
at a percentage response of 60.65%.

All continuous normal distributions have the property that between the
mean and one standard deviation 68% is included on either side, between the
mean and two standard deviations 95%, and between the mean and three standard
deviations 99.7% of the total area under the curve. Continuous normal distribution: applications in medical physics
The normal distribution is often used in radionuclide measurements and
imaging to fit to experimental data. In this case, the equation is modified as

P( x) = 100e

1 x x


where the maximum value of the distribution at x is normalized to 100.

The spatial resolution of imaging devices such as scintillation cameras and
positron emission tomography equipment is determined as the full width at half
maximum (FWHM) response of a normal distribution fitted to a point or line
spread function (Fig.5.6). The FWHM of the imaging device used in Fig.5.6



was 23.6mm. The relation for a normal distribution between the FWHM and
standard deviation can be derived by setting P(x)=50 and solving Eq.(5.36):
FWHM=2.355 (5.37)
For the imaging system used in Fig.5.6, the standard deviation =10mm.
The value of the response P(x) is 60.65% at a distance of = x (Eq.(5.36)).
The value of the standard deviation can, therefore, also be obtained from the
measured percentage response curve by finding the x value at a percentage
response of 60.65% (Fig.5.6).
In radionuclide energy spectroscopy, the photopeak distribution can be
fitted to a normal distribution (Fig.5.1). The energy resolution of scintillation
detectors is expressed as the FWHM of the photopeak distribution divided by
the photopeak energy E. The energy spectrum in medical physics applications
is measured in kiloelectronvolts or megaelectronvolts. The fractional energy
resolution RE is:
RE =

FWHM 2.355



5.4.1. Assumption
A valuable application of counting statistics applies to the case in which
only a single measurement of a particular quantity is available and the uncertainty
associated with that measurement is required. The square root of the sample
variance should be a measure of the deviation of any one measurement from
the true mean value and will serve as an index of the degree of precision that
should be associated with a measurement from that set.
As only a single measurement is available, the sample variance cannot be
calculated directly using Eqs(5.3) or (5.9) and must be estimated by analogy
with an appropriate statistical model. The appropriate theoretical distribution
can be matched to the available data if the measurement has been drawn from
a population whose theoretical distribution function is predicted by either a
Poisson or Gaussian distribution. As the value of the single measurement x is the



only information available, it is assumed that the mean of the distribution is equal
to the single measurement:


Having obtained an assumed value for x , the entire predicted probability

distribution function P(x) is defined for all values of x.
The expected sample variance s2 can be expressed in terms of the variance
of the selected statistical model:
s 2 = 2 = x x (5.40)

Therefore, the best estimate of the deviation from the true mean, which
should typify a single measurement x, is given by:
s = x (5.41)

To illustrate the application of Eq.(5.41), it is assumed that the probability

distribution function is Gaussian with a large value for the measurement x. The
range of values x or x x will contain the true mean with 68% probability.
If it is assumed that there is a single measurement x=100, then:
x = 100 = 10

In Table5.3, the various options available in quoting the uncertainty to be

associated with the single measurement are shown. The conventional choice is
to quote the measurement x plus orminus the standard deviation or 10010.
This interval is expected to contain the true mean x with a probability of 68%.
The probability that the true mean is included in the range can be increased by
expanding the interval associated with the measurement as is shown in Table5.3.
For example, to achieve a 99% probability that the true mean is included, the
interval must be expanded by 2.58. In the example, the range is then 10025.8.
When errors are reported, the associated probability level should be stated
in the report under methods.



Table5.3. Examples of Error Intervals for a Single

Measurement x=100

Probability that the true mean



















(relative )

x is included (%)

5.4.2. The importance of the fractional F as an indicator of the precision

of a single measurement in sample counting and imaging
The relation between the precision and a single counting measurement
x is given by Eq.(5.40). The precision, expressed as the standard deviation ,
will increase proportionally to the square root of the measurement x. Thus, if
the value of the single measurement x increases, the standard deviation will
also increase. The increase in the standard deviation will be smaller than that
of the measurement x. The relation between the standard deviation and the
single measurement is best demonstrated by calculating the relative or fractional
standard deviation F:
F =


Thus, the recorded number of counts or the value of the single measurement
x completely determines the relative precision. The relative precision decreases
as the number of counts increases. Therefore, to achieve a required relative
precision, aminimum number of counts must be accumulated.
The following example illustrates the important relation between the
relative precision and the number of counts recorded. If 100 counts are recorded,
the relative standard deviation is 10%. If 10 000 counts are recorded, the relative
standard deviation reduces to 1%. This example demonstrates the importance of
acquiring enough counts to meet the required precision.
It is easier to achieve the required precision when samples in counting tubes
are measured than when in vivo measurements on patients are performed. The


single measurement from a high count rate radioactive sample in a counting tube
will be obtained in a short time. However, if a low activity sample is measured,
the measurement time will have to be increased to achieve the desired precision.
The desired precision can be conveniently obtained by using automatic sample
counters. These counters can be set to stop counting after a preset time or preset
counts have been reached. By choosing the preset count option, the desired
precision can be achieved for each sample.
The acquisition time of in vivo measurements using collimated detector
probes, such as thyroid iodine uptake studies or imaging studies, can often not
be increased to achieve the desired precision as a result of patient movement. In
these single measurements, high sensitivity radiation detectors or a higher, but
acceptable radioactive dose, can be selected.
The precision of a single measurement is very important during radionuclide
imaging. If the number of counts acquired in a picture element or pixel is low, a
low precision is obtained. There will then be a wide range of fluctuations between
adjacent pixels. As a result of the poor quality of the images, it would only be
possible to identify large defect volumes or defects with a high contrast. To detect
a defect, the measured counts from the defect must lie outside the range of the
background measurement plus orminus two standard deviations (x2). During
imaging, the number of counts measured in a target volume will be determined
by the acquisition time, activity within the target volume and the sensitivity of
the measuring equipment. The sensitivity of imaging equipment can be increased
by increasing the FWHM spatial resolution. There is a trade-off between single
sample counting precision and the spatial resolution of the imaging device to
obtain images that would provide the maximum diagnostic value during visual
interpretation of the images by nuclear medicine physicians.
Counting statistics are also very important during image quantification
such as measuring renal function, left ventricular ejection fraction and tumour
uptake. During quantification, the accumulated counts by an organ or within a
target volume have to be accurately determined. In quantification studies, the
background activity, attenuation and scatter contributions have to be corrected.
These procedures further reduce the precision of quantification.
5.4.3. Caution on the use of the estimate of the precision of a
single measurement in sample counting and imaging
All conclusions are based on the measurement of a counted number of
success (number of heads in coin tossing). In nuclear measurements or imaging,
the estimate of the precision of a single measurement by using = x can only
be applied if x represents a counted number of success, that is the number of
events recorded in a given observation time.


The estimate of the precision of a single measurement by using = x

cannot be used if x is not a directly measured count. For example, the association
does not apply to:
Counting rates;
Sums or differences of counts;
Averages of independent counts;
Pixel counts following tomographic image reconstruction;
Any derived quantity.
In these cases, the quantity is calculated as a function of the number of
counts recorded. The error to be associated with that quantity must be calculated
according to the error propagation methods outlined in the next section.
The preceding section described methods for estimating random error or
the precision of a single measurement during nuclear measurements or imaging.
Most procedures in nuclear medicine involve multiple nuclear measurements
and imaging procedures for the calculation of results such as thyroid iodine
uptake, ejection fraction, renal clearance, blood volume or red cell survival time,
on which clinical diagnosis is based. Similarly, internal dosimetry is performed
using nuclear measurements and imaging data. To estimate the corresponding
precision in the derived quantity, how the error associated with the initial
measurements propagates through the calculations that were performed to arrive
at the required result has to be followed. This is done by applying the error of
propagation formulas. The variables used in the calculation of errors must be
independent to avoid effects of correlation. It is assumed that the error in nuclear
measurements arises only from random fluctuations in the decay rate and is
statistically independent of other errors.
The error of propagation formulas applies to measurements that are
obtained from a continuous distribution as well as to Poisson and discrete
normal distributions. The measurements from continuous distributions will be
represented by x1, x2, x3... with variances of (x1)2, (x2)2, (x3)2... These equations
can be used to estimate precision in measurements such as height and weight.
Discrete nuclear measurements with Poisson or normal distribution
are represented by N1, N2, N3... with variances of (N1)2, (N2)2, (N3)2... or
N1, N2, N3...



5.5.1. Sums and differences

The product xs of the sums or difference of a series of measurements with a
continuous normal distribution is given by:


The variance of xs is given by:

(x1x2x3)2=(x1)2 + (x2)2 + (x3)2


The standard deviation is given by:

( x 1 x 2 x 3 ...) = ( x 1 ) 2 + ( x 2 ) 2 + ( x 3 ) 2 ... (5.45)

The fractional standard deviation is given by:

F ( x 1 x 2 x 3 ...) =

( x 1 ) 2 + ( x 2 ) 2 + ( x 3 ) 2 ...
x 1 x 2 x 3 ...


For counting measurements or measurements with a Poisson or discreet

normal distribution, the variance is given by:
( N 1 N 2 N 3 ...) 2 = N 1 N 2 N 3 ... (5.47)

The standard deviation is given by:

( N 1 N 2 N 3 ...) = N 1 + N 2 + N 3 ... (5.48)

The fractional standard deviation is given by:

F ( N 1 N 2 N 3 ...) =

N 1 + N 2 + N 3 ...
N 1 N 2 N 3 ...


These equations apply to mixed combinations of sums and differences.



TABLE5.4. Uncertainty after Summing and Subtracting


N2 N1

N2 N1















N1 N2







N1 + N2







The influence on the standard deviation and fractional standard deviation

of summing and subtracting values N1 and N2 is demonstrated in Table5.4. The
following conclusions can be drawn:
The standard deviation for N1 N2 and N1 + N2 is the same for the same
values of N1 and N2, but the fractional standard deviation F is different;
The fractional standard deviation for differences is large when the
differences between the values are small.
This is the reason why it is important to limit the background to a value as
low as possible in counting procedures. In imaging, when scatter or background
correction is performed by subtraction, image quality deteriorates as a result of
the increased uncertainty in the pixel values.
5.5.2. Multiplication and division by a constant
We define:
xM=Ax (5.50)
where A is a constant.
M=Ax (5.51)



A x x

F =

For counting measurements or measurements with a Poisson or discreet

normal distribution, the following applies:
xM=AN (5.53)
M = A N (5.54)


F =

Similarly, if:

xD =

where B is also a constant:

M =


F =

x B x
B x

For counting measurements or measurements with a Poisson or discreet

normal distribution, the following apply:
xD =


M =





F =

It should be noted that multiplying (Eqs(5.52) and (5.55)) or dividing

(Eqs(5.58) and (5.61)) a value by a constant does not change the fractional
standard deviation.
5.5.3. Products and ratios
The uncertainty in the product or ratio of a series of measurements
x1, x2, x3... is expressed in terms of the fractional uncertainties in the individual
results, F(x1), F(x2), F(x3)...
The product xP of the products or ratios of a series of measurements with a
continuous normal distribution is given by:

x P = x1



... ... ... ... ... (5.62)

The notation means x 1 x 2 or x 1 x 2 . These equations apply to mixed

combinations of sums and differences.
The fractional variance of xP is given by:
F (x1



... ...) 2 = F ( x 1 ) 2 + F ( x 2 ) 2 + F ( x 3 ) 2 ... ... (5.63)

The fractional standard deviation is given by:

F (x1



... ...) = F ( x 1 ) 2 + F ( x 2 ) 2 + F ( x 3 ) 2 ... ... (5.64)

The standard deviation is given by:

( x 1



... ) = F ( x 1 ) 2 + F ( x 2 ) 2 + F ( x 3 ) 2 ... ( x 1



... ) (5.65)

For counting measurements or measurements with a Poisson or discreet

normal distribution, the product or ratio is given by:
NP = N1



... ... (5.66)

The fractional variance of NP is given by:

F (N 1




... ...) 2 =

+ ... ...
N1 N2 N3


The fractional standard deviation is given by:

F (N 1



... ...) =

+ ... ... (5.68)
N1 N2 N3

The standard deviation is given by:

( N 1



... ...) =

+ ... ... ( N 1
N1 N2 N3



... ...) (5.69)


5.6.1. Multiple independent counts Sum of multiple independent counts
If it is supposed that there are n repeated counts from the same source for
equal counting times and the results of the measurements are N1, N2, N3....... Nn
and their sum is Ns, then:
N s = N 1 + N 2 + N 3 ... (5.70)

According to the propagation of error for sums and Eq.(5.48):

N = N 1 + N 2 + N 3 ... = N s (5.71)

The results show that the standard deviation for the sum of all counts is the
same as if the measurement had been carried out by performing a single count,
extending over the period represented by all of the counts. Mean value of multiple independent counts
If the mean value N of the n independent counts referred to in the previous
section is calculated, then:




Equation (5.72) is an example of dividing an error-associated quantity N

by a constant n. Equation (5.51), therefore, applies and the standard deviation of
the mean or standard error is given by:
N =




It should be noted that the standard deviation for a single measurement Ni

(Eq.(5.41)) is N i = N i .
A typical count will not differ greatly from the mean N i N . Thus, the
mean value based on n independent counts will have an expected error that is
smaller by a factor of n compared with any single measurement on which the
mean is based. To improve the statistical precision of a given measurement by a
factor of two, the counting time must, therefore, be increased four times.
5.6.2. Standard deviation and relative standard
deviation for counting rates
If N counts are accumulated over time t, then the counting rate R is given



In the above equation, it is assumed that the time t is measured with a very
small uncertainty, so that t can be considered a constant. The calculation of the
uncertainty associated with the counting rate is an application of the propagation
of errors, multiplying by a constant (Eq.(5.60)):
R =


The fractional standard deviation is calculated using Eq.(5.61):

F =


The above equations illustrate the calculation of uncertainties if calculations

are required to obtain a value, and the equation for a single value (Section 5.3)
cannot be applied. The following example illustrates the use of the equations.


STATISTICS FOR RADIATION MEASUREMENT Example: comparison of error of count rates and counts accumulated

The activity of two samples is measured. Sample 1 is counted with a
counter that is set to stop when a count of 10000 is reached. It takes 100 s to
reach 10000 counts. Sample 2 is counted using an automatic sample changer.
The activity of the sample is given as 10000 counts per second (cps) and the
sample was counted for 100 s.
Calculating the counting error associated with the measurements of
samples 1 and 2:
Sample 1:
The counts acquired: N=10 000 counts
Standard deviation (Eq.(5.41)): s N = 100 counts
Fractional standard deviation (Eq.(5.42)): F = 0.01 = 1%
Sample 2:
The count rate: 10000 cps
10 000
= 10 cps
Standard deviation (Eq.(5.75)): R =

Fractional standard deviation (Eq.(5.76)):

F =

= 0.001 = 0.1%
10 000 100

Although the counts acquired for sample 1 and the count rate of sample 2
were numerically the same, the uncertainties associated with the measurements
were very different. When calculations on counts are performed, it must be
determined whether the value is a single value or whether it is a value that has
been obtained by calculation.
5.6.3. Effects of background counts
Background counts are those counts that do not originate from the sample
or target volume or are unwanted counts such as scatter. The background
counts during sample counting consist of electronic noise, detection of cosmic
rays, natural radioactivity in the detector, and down scatter radioactivity from
non-target radionuclides in the sample. During in vivo measurements, such
as measurement of thyroid iodine uptake or left ventricular ejection fraction,
radiation from non-target tissue will also contribute to background. Scattered


radiation from target as well as non-target tissue will influence quantification and
will be included in the background. To obtain the true net counts, the background
is subtracted from the gross counts accumulated. The uncertainty of the true
target counts can be calculated using Eqs(5.48) and (5.49), and the uncertainty of
true count rates can be calculated using Eqs(5.75) and (5.76).
If the background count is Nb, and the gross counts of the sample and
background is Ng, then the net sample count Ns is:
N s = N g N b (5.77)

The standard deviation for Ns counts is given by Eq.(5.48):

( N s ) = N g + N b (5.78)

The fractional standard deviation for Ns counts is given by Eq.(5.49):

F (N s ) =

Ng + Nb
Ng Nb


If the background count rate is Rb, acquired in time tb, and the gross count
rate of the sample and background is Rg, acquired in time tg, then the net sample
count rate Rs is:
Rs = Rg R b (5.80)

The standard deviation for a count rate Rs is given by Eqs (5.45) and(5.75):
(Rs ) =



The fractional standard deviation for a count rate Rs is given by Eqs (5.46)
F (Rs ) =



Rg R b


If the same counting time t is used for both sample and background
(R s ) =


Rg + R b



F (R s ) =

Rg + R b
t (R g R b )

(5.84) Example: error in net target counts following background correction

The following example illustrates the application to determine the
uncertainty in the measurement of target volume counts following background
correction. A planar image of the liver is acquired for the detection of tumours.
Two equal sized ROIs, ROI1 and ROI2, were selected to cover the areas of the
two potential tumours. The gross counts Ng in ROI1 were 484 counts (Table5.5)
and in ROI2 484 counts. The background counts Nb selected over normal tissue of
the same area as for the gross counts were 441 and 169 counts. How to calculate
the uncertainties in the tumor volume net counts is presented.
The difference and error associated with the difference (Eq.(5.77)
Eq. (5.79)) when Ng Nb are:
Ng Nb=484 441=43 counts
( N g N b ) = 484 + 441 = 30.4 counts
F (N g N b ) =

484 + 441
= 0.7073
484 441

P ( N g N b ) = 70.7%

The influence on the standard deviation and fractional standard deviation of

background correction for Ng Nb and Ng Nb is demonstrated in Table5.5. The
following conclusion can be drawn: the fractional F and percentage P standard
deviations significantly increase when the background increases relative to the
net counts.
This is the reason why it is important in measurements of radioactivity to
acquire as many counts as possible to decrease the uncertainty in detection of
target volume radioactivity. The following example illustrates the application
to determine the uncertainty in the measurement of target volume count rate
following background correction.




Ng Nb

Counts counts

Ng Nb

P (%) Source Counts counts

P (%)































3 (Ns)




3 (Ns)


Counts Significant Example: error in net target count rate following background correction
A planar image of the liver is acquired for the detection of tumours.
Two equal sized ROIs, ROI1 and ROI2, were selected to cover the areas of
the two potential tumours. The gross count rate Rg in ROI1 was 484 counts
perminute (cpm) (Table5.6) and in ROI2 484 cpm. The background count rates
Rb selected over normal tissue of the same area as for the gross counts were 441
and 169 cpm. The acquisition time of the image was 2min. How to calculate the
uncertainties in the tumor volume net counts is presented.
The difference and error associated with the difference (Eq.(5.80)
Eq.(5.82)) when Rg Rb are:
Rg R b = 484 441 = 43 cpm

(R g R b ) =

484 441
= 21.5 cpm

484 441
2 = 0.5001
F (R g R b ) =
484 441
P (Rg R b ) = 50.0%

The influence on the standard deviation and fractional standard deviation

of background correction for Rg Rb and Rg Rg is demonstrated in Table5.6.


Again, it is shown that the fractional standard deviation F significantly increases

when the background count rate increases relative to the net target count rate.
TABLE5.6. Calculation of Uncertainties in Count Rates as A
Result of Background Correction
Rg Rb

Count rate

P (%)


Rg Rb

Count rate
F (cpm)

P (%)

































3 (Rs)



3 (Rs)



5.6.4. Significance of differences between counting measurements

If N1 and N2 counts are measured in two counting measurements, the
difference (N1 N2) between the measured counts may be a result of random
variations in the counting rate or may be as a result of an actual difference.
The statistical significance of the difference is evaluated by comparing it to the
expected random error expressed as the standard deviation d of the difference.
If (N1 N2) > 2(N1 N2), there is a 5% chance that the difference is caused by
random error (seeTable5.3). If:
N 1 N 2 > 3( N 1 N 2 ) (5.85)

there is a 0.3% chance that the difference is caused by random error and this
difference is considered significant.
The examples in the previous section to determine whether tumours
were present following a liver scan illustrate the application to determine the
significance of the difference between two counts (Table5.5). The net counts
and uncertainty over two tumour areas were calculated. Do the counts over the
tumour areas significantly differ from the normal background area?
For the difference for Ng Nb (Table5.5) to be significant, Eq.(5.85) must


The difference of 43 cpm was less than the norm of 3(N1 N2) and the
difference is, therefore, not significant. It can be concluded with a smaller than
0.3% chance that there is not a tumour present.
An example when Ng Nb is also given in Table5.5. In this case, the
315 cpm counts difference was larger than 3(N1 N2) of 77 cpm. The difference
in this case is significant. It can be concluded with a smaller than 0.3% chance
that there is a tumour present.
The significance of differences between the counting rates of samples can
also be calculated. Two counting rates, R1 and R2, are acquired using counting
times t1 and t2.
The uncertainty associated with the difference is given by applying
Eqs(5.45) and (5.75):
( R1 R 2 ) =

R1 R 2

For the difference R1 R2 to be significant:

R1 R2 > 3(R1 R2) (5.87)
The examples in the previous section (Table5.6) to determine whether
tumours were present following a liver scan illustrate an application to determine
the significance of the difference between two count rates. The net count rate and
uncertainty over two tumour areas were calculated. Do the count rates over the
tumour areas significantly differ from the normal background area?
For the difference for Rg Rb (Table5.6) to be significant, Eq.(5.87) must
The difference count rate of 43 cpm was less than the 65 cpm which is the
norm of 3(R1 R 2 ) and the difference is, therefore, not significant. It can be
concluded with a smaller than 0.3% chance that there is not a tumour present.
An example when Rg Rb is also given in Table5.6. In this case, the
difference of 315 cpm was larger than 3(R1 R 2 ) which was 54 cpm. The
difference in this case is significant. It can be concluded with a smaller than 0.3%
chance that there is a tumour present.
5.6.5. Minimum detectable counts, count rate and activity
According to Eq.(5.85), if the difference of two measurements is larger
than three standard deviations, the difference is considered significant. Therefore,



theminimum net counts Nm that can be detected with 0.3% confidence is given
Nm=N1 N2=3(N1 N2) (5.88)
Nm=Ng Nb=3(Ng Nb) (5.89)
Solving this equation for Ng will give theminimum detectable gross
counts Nm:
(2 N b + 9) + 72 N b + 81

Ng =

(5. 90)

An approximation can be used by assuming that Ng Nb and:

N g N b + 3 2 N b (5.91)

Theminimum detectable activity Am can be calculated:

Am =


S is the sensitivity of the detection system usually expressed as count rate per
and t is the time that the background was counted. Example: calculation ofminimum activity that can be detected
A detector is to be used to detect 131I in the thyroid of radiation workers.
The background count was 441 counts measured over a period of 5min. The
acquisition time for the thyroid was also 5min. The sensitivity of the counter was
0.1 counts s1 Bq1. What is theminimum activity that can be detected?
From Eq.(5.90):
Ng =

(2 N b + 9) + 72 N b + 81

(2 441 + 9) + 72 441 + 81
= 535 counts



It should be noted that Ng Nb=94 counts and 3(Ng Nb)=94 as was

specified in Eq.(5.85). Theminimum detectable radioactivity is:
Am =

(535 441)
= 3.124 Bq
5 60 0.1

Theminimum detectable net count rate Rm is given by Eq.(5.89):

Rm = Rg R b > 3(Rg R b ) (5.93)

Solving this equation for Rg gives theminimum detectable gross count

rate Rm:

36 R b 81 36 R b

+ 2+
2 R b + +
t g

Rg =



An approximation can be used by assuming that Rg Rb and from Eqs(5.86)

and (5.87):
Rg R b + 3

Rb Rb

Theminimum detectable activity Am can be calculated:

Am =


where S is the sensitivity of the detection system usually expressed as count rate
per becquerel. Example 2: calculation ofminimum activity that can be detected
A detector is to be used to detect 131I in the thyroid of radiation workers.
The background count rate was 441 cpm measured over a period of 5min and the
thyroid count rate was measured over 1min. The sensitivity of the counter was
0.1 counts s1 Bq1. What is theminimum activity that can be detected?
From Eq.5.94:

2 441 + 9 + 36 441 + 81 + 36 441

Rg =
= 515 cpm



It should be noted that Rg Rb=74 cpm and 3(Rg Rb)=74 cpm as was
specified in Eq.(5.93). Theminimum detectable radioactivity is:
Am =

(515 441)
= 12.28 Bq
0.1 60

5.6.6. Comparing counting systems

It was concluded in Section 5.3.1 that a large number of counts have smaller
uncertainties expressed as the fractional standard deviation. In Section 5.6.3, it
was shown that if background counts increase, the uncertainty of the net counts
expressed as fractional standard deviation rapidly increases. Thus, it is desirable
to use a counting system with a high sensitivity and low background. However,
when the detector sensitivity is increased, the system will also be more sensitive
to background. The trade-off between sensitivity and background can be analysed
as follows.
It is considered that results from systems 1 and 2 are compared. The
acquisition times for gross and background counts are acquired over the same
time. From Eq.(5.79):
N g1 + N b1

F1( N S1 ) =

N g1 N b1

F2 ( N S2 ) =

N g2 + N b2
N g2 N b2

The fractional uncertainties for the net sample counts obtained with the two
systems are, therefore:
N g1 + N b1
N g1 N b1
F1( N S1 )
F2 ( N S2 )
N g2 + N b2
N g2 N b2


F1( N S1 )
< 1, then system 1 is statistically the preferred system. If
F2 ( N S2 )

F1( N S1 )
> 1 , then system 2 is preferred.
F2 ( N S2 )



Systems can be compared using the count rate and fractional standard
deviation for the count rate RS (Eq.(5.82)). To compare systems 1 and 2, the ratio
of the fractional standard deviation is calculated:
t g1

R b1
t b1

Rg1 R b1
F1(RS1 )
F2 (RS2 )
Rg2 R b2
t g2
t b2
Rg2 R b2

Equation (5.98) can be used to compare different counting times in the

same system for measuring fixed geometry samples. However, to obtain the
best energy window selection in a system, or to compare two systems, the same
counting time t should be used:
Rg1 + R b1
Rg1 R b1
F1(RS1 )
F2 (RS2 )
Rg2 + R b2
Rg2 R b2

It should be noted that Eqs(5.98) and (5.99) are the same except that in
Eq.(5.99) counts are substituted by counting rates.
Equation (5.99) can also be used in planar imaging. Different collimators
can be evaluated by comparing counts from a target region to a non-target or
background region. However, in imaging, spatial resolution is also important and
must be considered.
5.6.7. Estimating required counting times
It is supposed that it is desired to determine the net sample or target
counting rate Rs to within a certain fractional uncertainty F(Rs). It is supposed
further that the approximate gross sample Rga and background Rba counting rates
are known from preliminary measurements. If a counting time t is to be used
for both the sample or target and the background counting measurements, then
the time required to achieve the desired level of statistical reliability is given by




Rga + R ba
[ F (Rs )](Rga R ba ) 2 Example: calculation of required counting time

The counting time for a thyroid uptake study using a collimated detector is
to be determined. The preliminary measurement of the gross thyroid count rate is
Rga=900 cpm and background count rate Rba=100 cpm. What counting time is
required to determine the net count rate to within 5%?
Rsa=900 100=800 cpm

(900 + 100)
= 0.625 min
(0.05) (900 100)
(0.05) 2 (800) 2

The time for both the thyroid and background counts is 0.625min, resulting
in a total time of 1.25min.
5.6.8. Calculating uncertainties in the measurement
of plasma volume in patients
A plasma volume (PV) measurement is required on a patient and the
uncertainty in the PV measurement is to be calculated. The PV is measured by
using the dilution principle. A labelled plasma sample of a known volume is
prepared for injection into the patient. A standard sample with the same activity
and volume is also prepared for counting. The standard sample is diluted before
a sample is counted. Tenminutes after injection of the sample, a blood sample
is obtained, the plasma separated from the blood and the blood sample counted.
The PV is calculated using the following equation:
PV =

VD (5.100)

Net count rate per millilitre of standard sample Rs=Rs+b Rb;
Rb is the count rate of background;
Rs+b is the gross count rate per millilitre of standard sample;
Net count rate per millilitre of plasma sample Rp=Rp+b Rb;



Rp+b is gross count rate per millilitre of plasma sample;

V is volume of standard sample in millilitres with percentage uncertainty P(V);
and D is dilution of standard sample for counting with percentage
uncertainty P(D).


Uncertainty in values











17.89 0.559






4.472 2.236





((Rs+b )) 2 + ((R b )) 2

18.44 0.615



















R p

R p+b

10.95 0.913

((R p+b )) 2 + ((R b )) 2


(R ) 2 (R p )

R + R


0.040 1.333
0.150 3.000

R p

(R / R ) 2 (V ) 2


+ V
Rs / R p


0.492 3.283
6.000 3.000




((R / R )V)
(D) 2


(Rs / R p )V

The following values were used and measured:

Counting time t =10min
=53% mL
=3200 cpm
=1200 cpm
=200 cpm

11.83 1.183




The uncertainties are calculated step by step by applying the propagation of

errors principle (seeTable5.7)
The measured PV is, therefore, 3000133 mL or 30004.447%. It
should be noted that the uncertainty is expressed as one standard deviation. A
spreadsheet can be used efficiently to do the calculations in the above table. With
a spreadsheet, the influence in changing the counting time or uncertainties in
the measurement of the dilution and volume of the standard can be investigated.
These spreadsheets are ideally suited for calculations of uncertainties in routine
clinical investigations.
5.7.1. Energy resolution of scintillation detectors
We have directed our attention in the previous sections to determine the
uncertainty associated with the number of counts measured in a radioactive
sample or number of counts in an image pixel. Poisson statistics also play an
important role in other aspects of the detection of radiation. A statistical process
determines the energy resolution of a detector or the uncertainty associated
with the energy measurement of a detected photon. This is the reason why the
energy resolution of a solid state detector is significantly better than that of a
scintillation detector. The type of detector and the energy of the detected photons
determine the energy resolution or uncertainty in the energy of a detected photon.
The energy resolution for a detector system and a specific radionuclide does not
change from sample to sample. This is different from counting statistics where
the uncertainty is determined by the number of counts accumulated during a
measurement. Therefore, even for the same sample and same detector system, the
uncertainty can change if measurements are repeated following the decay of the
Another important consequence of statistics is that in scintillation cameras
the location of the position of incoming photons is based on the pulses detected
by the detectors. Therefore, the statistics of the detector system limits the spatial
resolution that can be achieved with an imaging device. A clear understanding of
the statistics associated with the detector when detecting a photon is, therefore,
In this section, we will investigate the statistical processes in scintillation
detectors, since they are widely used in nuclear medicine for sample counting
and imaging.



The operation of scintillation detectors can be considered a three stage

(a) The number x of light photons produced in the scintillator by the detected
(b) The fraction p of the light photons that will eject electrons from the
photocathode of the PMT;
(c) The multiplication M of these electrons multiplied at successive dynodes
before being collected at the anode.
The average number Ne of electrons produced at the anode is given by:
N e = xpM (5.101)

The fractional variance F2 in the electron number N for a three stage

cascade process is given by Eq.(5.63):
F2 ( N e ) = F2 ( x) + F2 ( px) + F2 ( pxM ) (5.102)

It can be shown that for dynodes with identical multiplication

. It is assumed that the production of light photons follows
M 1
a Poisson distribution and, therefore, F2 ( x) = . The fractional variance of
F2 (M ) =

the production of electrons from light photons at the photocathode is given by

Eq.(5.20) as F2 ( p) =
F2 ( N e ) =

1 p

1 1 1 p
x x p
xp (M 1)

The fractional energy resolution RE of detectors is expressed as the FWHM

divided by the mean photon energy (Section From Eq.(5.38):
RE =

FWHM 2.355(E) 2.355( N e )


From Eqs(5.103) and (5.104):

R E = 2.355



1 p

x 1 +
p(M 1)


5.7.2. Intervals between successive events

The time intervals separating random events are of interest in nuclear
measurements. Such an application is the calculation and measurement of the
paralysable dead time of counting systems.
If r is the average rate at which events are occurring, it follows that r dt is
the differential probability that an event will take place in the differential time
increment dt. For a radiation detector with unity efficiency, the time interval for
counting a single radionuclide is given by:

= lN


is the number of radioactive nuclei;

and is their decay constant.

In order to derive a distribution function to describe the time interval
between adjacent random events, it is first assumed that an event has occurred at
time t=0. What is the differential probability that the next event will take place
within a differential time dt after a time interval t?
Two independent processes must take place: no events may occur within
the time interval from 0 to t, but an event must take place in the next differential
time increment dt. The overall probability will then be given by the product of
the probabilities characterizing the two processes, or:
Probability of next event Probability of number
taking place in dt after
= of events during time
delay of t
from 0 to t
P1(t ) dt

= P(0)

Probability of
event during dt

r dt


The first factor on the right hand side follows directly from the earlier
discussion of the Poisson distribution. We seek the possibility that no events will
be recorded over an interval of length t for which the average number of recorded
events should be rt. From Eq.(5.23):
P(0) =

(rt ) 0 e rt



P(0) = e rt (5.107)

Substituting Eq.(5.107) into Eq.(5.106):

P1(t ) dt = re rt dt (5.108)

P1(t) is now the distribution function for intervals between adjacent random
events. Figure5.7 shows the simple exponential shape of the distribution.

FIG.5.7. Distribution for intervals between adjacent random events.

It should be noted that the most probable distribution is zero. The average
interval length is calculated by applying Eq.(5.8):

t =

P (t ) d t

tP1(t ) dt

te rt dt





5.7.3. Paralysable dead time

In the paralysable dead time model, a fixed dead time follows each event
during the live period of the detector. However, events that occur during the


dead period, although not recorded, still create another fixed dead time on the
system following the lost event. The recorded rate of events m is identical to the
rate of occurrences of time intervals between true events, which exceed . The
probability of intervals larger than can be obtained by integrating Eq.(5.108):
P2 (t ) dt =

P1(t ) dt = e r (5.110)

The rate occurrence m of such intervals is obtained by multiplying

Eq.(5.110) by the true rate r:
m = re r


There is no explicit solution for r; it must be solved iteratively to calculate r

from measurements of m and . This can be done using a spreadsheet.
BUSHBERG, J.T., SEIBERT, J.A., LEIDHOLDT, E.M., BOONE, J.M., The Essential Physics
of Medical Imaging, Lippincott Williams and Wilkins, London (2002).
CHERRY, S.R., SORENSON, J.A., PHELPS, M.E., Physics in Nuclear Medicine, Saunders,
Los Angeles, CA (2003).
DELANEY, C.F.G., FINCH, E.C., Radiation Detectors, Clarendon Press, Oxford (1992).
KNOLL, G.F., Radiation Detection and Measurement, John Wiley and Sons, New York (1989).
NU 1-2007, Performance Measurements of Gamma Cameras (2007).


Faculty of Applied Sciences,
Delft University of Technology,
Delft, Netherlands
6.1.1. Radiation detectors complexity and relevance
Radiation detectors are of paramount importance in nuclear medicine. The
detectors provide a wide range of information including the radiation dose of
a laboratory worker and the positron emission tomography (PET) image of a
patient. Consequently, detectors with strongly differing specifications are used.
In this chapter, general aspects of detectors are discussed.
6.1.2. Interaction mechanisms, signal formation and detector type
A radiation detector is a sensor that upon interaction with radiation
produces a signal that can preferably be processed electronically to give the
requested information. The interaction mechanisms for X rays and rays are the
photoelectric effect, Compton scattering and pair formation, where the relative
importance depends on the radiation energy and the interaction medium. These
processes result in the production of energetic electrons which eventually
transfer their energy to the interaction medium by ionization and excitation.
Charged particles, such as particles, transfer their energy directly by ionization
and excitation. In all cases, the ionization results either in the production of
charge carriers, viz. electrons and ions in a gaseous detection medium, and
electrons and holes in a semiconductor detector material, or in the emission of
light quanta in a scintillator. These processes represent the three major groups
of radiation detectors, i.e. gas filled, semiconductor and scintillation detectors.
In the former two cases, a signal, charge or current is obtained from the
detector as a consequence of the motion of charge in the applied electric field
(Figs6.1(a)and (b)). In the scintillation detector, light emission is observed by
means of a light sensor that produces observable charge or current (Fig.6.1(c)). A
detailed discussion is presented in Sections 6.26.4.


6.1.3. Counting, current, integrating mode

In radiology and radiotherapy, radiation detectors are operated in current
mode. The intensities are too high for individual counting of events. In nuclear
medicine, on the contrary, counting mode is primarily used. Observing individual
events has the advantage that energy and arrival time information are obtained,
which would be lost in current mode. In the case of a personal dosimeter, the
detector is used in integrating mode. The dose is, for example, measured monthly.
Furthermore, instead of real time observation, the information is extracted at a
much later time after the actual interaction.



Light quanta



Light sensor

FIG.6.1. Principle of operation of (a) a gas filled detector, i.e. an ionization chamber;
(b)a semiconductor detector, i.e. a silicon detector; and (c)a scintillation detector. The former
two detectors are capacitors. The motion of charge results in an observable signal. The light of
a scintillation detector is usually detected by a photomultiplier tube.

6.1.4. Detector requirements

The quality of a radiation detector is expressed in terms of sensitivity,
energy, time and position resolution, and the counting rate a detector can handle.
Obviously, other aspects such as cost, machinability and reliability are also very
important. The latter will not be discussed in this chapter. Sensitivity
In radiation detection, the sensitivity depends on (i) the solid angle
subtended by the detector and (ii) the efficiency of the detector for interaction
with the radiation. The first point will be obvious and is not discussed further.
In nuclear medicine, relevant X ray and ray energies are in the range of
~30511keV. The detection efficiency is governed by the photoelectric effect
and Compton scattering only. The attenuation length (in centimetres) of the


former is proportional to Zeff34, where is the density and Zeff is the effective
atomic number of the compound. Compton scattering is almost independent of Z;
it is just proportional to . The density of a gas filled detector is three orders of
magnitude smaller than that of a solid state detector. Thus, solid state detectors
are very important in nuclear medicine. At 511keV, even the highest possible
and Zeff are needed. Gas filled detectors are used in dosimetry. Energy, time and position resolution
Energy, time and position resolution depend on a number of factors. These
are different depending on the physical property considered and the type of
detector; yet, there is one aspect in common. Resolution is strongly coupled to
the statistics of the number of information carriers. For radiation energy E, this
number is given by N=E/W in which W is the mean energy needed to produce
an information carrier. Typical W values are shown in Table6.1. As the smallest
number of information carriers in the process of signal formation is determinative,
for scintillation the effect of the light sensor is also shown. From the W values,
it can be seen that semiconductor detectors produce the largest number of
information carriers and inorganic scintillators coupled to a photomultiplier tube
(PMT) the smallest. If a ray energy spectrum is measured, the observed energy
resolution is defined as the width of a line at half height (FWHM: full width at
half maximum) E divided by its energy E. With N =E/W, and N being the
corresponding FWHM:
= 2.35

N is 2.35 for a Gaussian distribution;
2 is FN for the variance;
and F is the Fano factor. For gas filled detectors, F=0.050.20, for
semiconductors F 0.12. For a scintillator, F=1.
Using the corresponding F and W values, it can be seen from Eq.(6.1)
that the energy resolution of a semiconductor is ~16 times higher than that
of an inorganic scintillator PMT. In this discussion, other contributions to the
energy resolution were neglected, viz. from electronic noise in the case of the
semiconductor detector and from scintillator and PMT related effects in the



other case. Nevertheless, the large difference, by an order of magnitude, is

characteristic of the energy resolutions.



Detector type
Gas filled (electronion)
Semiconductor (electronhole)
Inorganic scintillator (light quantum)
Inorganic scintillator + photomultiplier tube (electron)
Inorganic scintillator + silicon diode (electronhole pair)


W (eV)

In nuclear medicine, time resolution is mainly of importance for PET. Time

resolution depends primarily on two factors, the rise time and the height of the
signal pulses. The effect of the former can be understood by considering that it
is easier to measure the position of a pulse on a timescale with an accuracy of
100 ps if the rise time is 1 ns, than if it is 10 ns. The pulse height is important
because there is noise as well. The higher the pulse relative to the noise, the
easier it is to determine its position. In addition, time jitter due to pulse height
(energy) variation will become less important. If time resolution is the issue, the
fast response and fast rise time of inorganic scintillators and the fast response of
the light sensors make the scintillator the preferred detector.
Position resolution can be obtained easiest by pixelating the detector at a
pitch corresponding to the requested resolution. In nuclear medicine, position
resolution is an issue in ray detection in the gamma camera and in single photon
emission computed tomography (SPECT) and PET detection systems. In the
latter, pixelated scintillators are used and the position resolution of a detector
is determined by the pitch. More recently, studies have been published on the
use of monolithic scintillator blocks in PET. Light detection occurs by means
of pixelated sensors. In principle, this is analogous to the gamma camera. A
relatively broad light distribution is measured using pixels that are smaller in size
to define the centre of the distribution, thus obtaining a position resolution that is
even better than the pixel size. Counting rate and dead time
An achievable counting rate depends on (i) the response time of a detector,
i.e. the time it takes to transport the charge carriers to form the signal or to emit


the scintillation light, and (ii) the time needed to process the signals and to handle
the data. For a better understanding, the concept of dead time is introduced.
It is theminimum time separation between interactions (true events) at which
these are counted separately. Non-paralysable and paralysable dead time are
considered. In the former case, if within a period of time after a true event a
second true event occurs, it cannot be observed. If the second event occurs at a
time t > , it will be counted. The dead period is of fixed length . Defining true
event rate T (number per unit time) and counting rate R, the fraction of time the
system is dead is given by R and the rate of loss of events is TR. Considering
that the latter is also T R, the non-paralysable case can be derived:


If in the paralysable model a second event occurs at t > after the first
event, it will be counted. If a second event occurs at t < after the first event, it
will not be counted. However, in the paralysable case, if t < , the second event
will extend the dead time with a period from the moment of its interaction. If a
third event occurs at t > after the first event but within a period of time after
the second event, it will not be counted either. It will add another period of . The
dead time is not of fixed length. It can become much larger than the basic period
and in this case it is referred to as extendable dead time. Only if an event
occurs at time > after the previous event will it be counted. In this case, the
counting rate is the rate of occurrences of time intervals > between events, for
which the following can be derived:
R =Te T (6.3)

Figure6.2 demonstrates the relation between R and T for the two cases
above and for the case of =0, i.e. R=T.
6.2.1. Basic principles
The mode of operation of a gas filled detector depends strongly on the
applied voltage. In Fig.6.3(a), the signal amplitude is shown as a function of the
voltage V. If upon interaction with radiation an energetic electron ploughs through
the gas, the secondary electrons produced will tend to drift to the anode and the
ions to the cathode (seeFig.6.1(a)). If the voltage is relatively low, the electric



FIG.6.2. Counting rate R as a function of true event rate T in the absence of dead time
(R=T), in the non-paralysable case and in the paralysable case.

field E is too weak to efficiently separate the negative and positive charges. A
number of them will recombine. The full signal is not observed this is in the
recombination region. Increasing the voltage, more and more electrons and ions
escape from recombination. The region of full ionization is now reached. For
heavier charged particles and at higher rates, this will happen at a higher voltage.
The signal will become constant over a wide voltage range. Typical operating
voltages of an ionization chamber are in the range of 5001000 V.
For the discussion of operation at stronger electric fields, cylindrical
detector geometry with a thin anode wire in the centre and a metal cylinder as
cathode (seeFig.6.3(b)) is introduced. The electric field E(r) is proportional to the
applied voltageV and inversely proportional to the radius r. At a certain voltage
VT, the threshold voltage, the electric field near the anode wire is so strong that a
drifting electron will gain enough energy to ionize a gas atom in a collision. The
proportional region is entered. If the voltage is further increased, the ionization
zone will expand and an avalanche and significant gas amplification are obtained.
At normal temperature and pressure, the threshold electric field ET 106 V/m. For
parallel plate geometry with a depth of ~1cm, this would imply that VT 10kV,
which is not practicable. Due to the r1 dependence, in the cylindrical geometry,
manageable voltages can be applied for proportional operation (13kV). As long
as the gas gain M is not too high (M 104), it is independent of the deposited
energy. This is referred to as the proportional region and proportional counter.
If the voltage is further increased, space charge effects will start to reduce the
effective electric field and, consequently, affect the gain. This process will start
at a lower voltage for the higher primary ionization density events. The limited
proportionality region is entered. With further increasing voltage, the pulse


height will eventually become independent of the deposited energy. This is the
GeigerMller region.

Pulse height (log scale)











Cylinder geometry

High voltage

FIG.6.3. (a) Pulse height as a function of applied high voltage for gas filled detectors;
(b)cylindrical detector geometry.

Instead of one wire in a cylindrical geometry, many equidistant parallel

anode wires at a pitch of 12mm can be positioned in a plane inside a box with
the walls as cathode planes. This multiwire proportional chamber (MWPC) is
employed in autoradiography. The technique of photo-lithography made it
possible to introduce micro-patterned detectors that operate analogously to
the MWPC. Examples are the micro-strip gas chamber and the gas electron
multiplier. Spatial resolutions are of the order of 0.1mm.
6.3.1. Basic principles
As shown in Fig.6.1(b), a semiconductor detector is a capacitor. If upon
interaction with radiation, electrons are lifted from the valence band into the
conduction band, the transport of the charge carriers in an applied electric field is
observed. However, if a voltage difference is supplied to electrodes on opposite
sides of a slab of semiconductor material, in general, too high a current will flow
for practical use as a detector. At room temperature, electrons are lifted from the
valence band into the conduction band by thermal excitation due to the small gap
(Egap 1 eV). The resulting free electrons and holes cause the current. A solution
is found in making a diode of the semiconductor, operated in reverse bias. Silicon


is used as an example. The diode structure is realized by means of semiconductorelectronics technology. Silicon doped with electron-donor impurities, called
n-type silicon, can be used to reduce the number of holes. Electrons are the
majority charge carriers. Silicon with electron-acceptor impurities is called p-type
silicon; the number of free electrons is strongly reduced. The majority charge
carriers are the holes. When a piece of n-type material is brought into contact
with a piece of p-type material, a junction diode is formed. At the junction,
a space charge zone results, called a depletion region, due to diffusion of the
majority charge carriers. When a positive voltage is applied on the n-type silicon
side with respect to the p-type side, the diode is reverse-biased and the thickness
of the depletion layer is increased. If the voltage is high enough, the silicon will
be fully depleted. There are no free charge carriers left and there is virtually no
current flowing. Only a small current will remain, the leakage or dark current.
To make a diode, n-type silicon is the starting material and a narrow zone
is doped with impurities to make a p+n junction, as indicated at the bottom of
Fig.6.1(b). The notation p+ refers to a high doping concentration. For further
reduction of the leakage current, high purity silicon and a blocking contact are
used, i.e. an n+ doping at the n-type side, also indicated in Fig.6.1(b). If the
leakage current is still problematic, the temperature can be decreased. The use
of high purity semiconductor material is not only important for reducing the
leakage current. Energy levels in the gap may trap charge carriers resulting from
the interaction with radiation and the energy resolution of a detector would be
The above described approach is not the only way to make a detector.
It is possible to start with p-type material and make an n+p junction diode.
Furthermore, it is possible to apply a combination of surface oxidation and
deposition of a thin metal layer. Such contacts are called surface barrier contacts.
If the thickness of a detector is <1mm, it is even possible to use intrinsic
silicon, symbol i, with p+ and n+ blocking contacts on opposite sides (pin
configuration). For thicker silicon detectors, yet another method is used. In
slightly p-type intrinsic silicon, impurities are compensated for by introducing
interstitial Li ions that act as electron donors. The Li ions can be drifted over
distances of ~10mm. Furthermore, if the bandgap of a semiconductor is large
enough, just metal contacts will suffice.
Important parameters are the mobilities, e and h, and the lifetimes, e and
h, of electrons and holes, respectively. The drift velocity e,h in an electric field E
is given by the product of the mobility and the field strength. Consequently, for
a given detector size and electric field, the mobilities provide the drift times of
the charge carriers and the signal formation times. From the mobilities and the
lifetimes, information on the probability that the charge carriers will arrive at the



collecting electrodes is obtained. The path length a charge carrier can travel in its
lifetime is given by:
e,h e,h = e,h e,h E (6.4)

If this is not significantly longer than the detector depth, charge carriers
will be lost.
6.3.2. Semiconductor detectors
Some properties of semiconductor detector materials of relevance for
nuclear medicine, viz. the density , effective atomic number for photoelectric
effect Zeff, Egap and W value, the mobilities e,h and the products of the mobilities,
and the lifetimes of the charge carriers, are presented in Table6.2.
Silicon is primarily of interest for (position sensitive) detection of
low energy X rays, particles and light quanta. The latter are discussed in

Si (300 K)





Si (77 K)






















Ge (77 K)









CdTe (300 K)









(CZT-300 K)









HgI2 (300 K)








See Section

For X ray detection in the range of ~300 eV to 60keV, planar circular

Li drifted pin detectors notated Si(Li) are commercially available with a
thickness up to 5mm. Diameters are in the range of 420mm. For typical field
strengths of ~1000 V/cm, the drift times to the electrodes are on the order of
tens of nanoseconds. Energy resolutions (FWHM) at 5.9keV are ~130220 eV if


operated at 77 K. Position sensitive silicon detectors with a large variety of pixel

structures are commercially available. Silicon detectors are also used in personal
Germanium, with its higher density and atomic number, is the basic
material for high resolution ray spectroscopy. Detectors are made of high purity
material. Large volume detectors are made of cylindrical crystals with their
core removed (coaxial geometry). High purity n-type or p-type is used with the
corresponding junction contacts on the outside and the blocking contacts on the
inside. Germanium detectors are operated at 77 K. Cylindrical detectors up to
a diameter of ~10cm and a height of ~10cm are commercially available. Drift
times to the electrodes can be as large as ~100 ns. Typical energy resolutions are
~1keV at 122keV ray energy and ~2keV at 1332keV.
Cadmium telluride (CdTe) and cadmium zinc telluride (CZT) are of interest
because their atomic number is significantly higher than that of germanium,
and room temperature operation is possible due to the larger bandgap. High
purity n-type or p-type material is used. The energy resolution is worse than
that of Ge detectors, e.g. 2.5% FWHM at 662keV. This is primarily due to the
relatively short lifetime of the holes, resulting in incomplete charge collection.
Electronic correction techniques are used and/or detectors with special electrode
configurations (small pixels or grids) are made to observe the electron signal
only. Detector dimensions are up to approximately 25mm25mm10mm.
Detectors of 25mm25mm5mm with 16 pixels16 pixels are available.
These detectors are used, for example, for innovation of SPECT.
In principle, HgI2 (mercury iodide) is an attractive material for efficient
ray detection because of the large density and high atomic number. Owing to
the relatively large bandgap, room temperature operation is possible. However,
the mobilities are low and charge collection, in particular of the holes, is
poor. Consequently, application is limited to detector thicknesses 10 mm.
Field strengths of 2500 V/cm are applied and analogous to CdTe and CZT,
methods are used to observe the electron signal only. Detector areas are up to
6.4.1. Basic principles
Scintillation of a material is the prompt emission of light upon interaction
with radiation. In nuclear medicine, inorganic ionic crystals are most important.
They combine high density and atomic number with a fast response and a high
light yield, and large crystals can be grown. These crystals form the backbone for


X ray and ray detection. Another group is formed by organic scintillators, viz.
crystals, plastics and liquids, which have a low density and atomic number, and
are primarily of interest for counting of particles. In some inorganic scintillator
materials, metastable states (traps) are created that may live from milliseconds to
months. These materials are called storage phosphors. Scintillators and storage
phosphors are discussed later in this section. However, as light detection is of
paramount importance, light sensors are introduced first.
6.4.2. Light sensors Photomultiplier tubes
The schematic of a scintillation detector is shown in Fig.6.4(a). A
scintillation crystal is coupled to a PMT. The inside of the entrance window of
the evacuated glass envelope is covered with a photocathode which converts
photons into electrons. The photocathode consists of a thin layer of alkali
materials with very low work functions, e.g. bialkali K2CsSb, multialkali
Na2KSb:Cs or a negative electron affinity (NEA) material such as GaAs:Cs,O.
The conversion efficiency of the photocathode , called quantum efficiency,
is strongly wavelength dependent (seeFig.6.5). At 400 nm, =2540%. The
emitted electrons are focused onto the first dynode by means of an electrode
structure. The applied voltage is in the range of 200500 V, and the collection
efficiency 95%. Typical dynode materials are BeOCu, Cs3Sb and GaP:Cs.
The latter is an NEA material. If an electron hits the dynode, electrons are
released by secondary emission. These electrons are focused onto the next
dynode and secondary electrons are emitted, etc. The number of dynodes n is in
the range of 812. The signal is obtained from the last electrode, the anode. At
an inter-dynode voltage of ~100 V, the multiplication factor per dynode 5. In
general, a higher multiplication factor is applied for the first dynode, e.g. 1 10,
to improve the single-electron pulse resolution, and consequently the signal to
noise ratio. Starting with N photons in the scintillator and assuming full light
collection on the photocathode, the number of electrons Nel at the anode is given
N el = 1 n1 N (6.5)

Gains of 106107 are obtained. A negative high voltage (10002000 V) is

often used with the anode at ground potential and care must be taken of metal
parts near the cathode. Furthermore, the detector housing should never be opened
with the voltage on. Exposure to daylight would damage the photocathode




Window PMT


Window PMT

Microchannel plates




Optical coupling

Focusing electrodes

Photocathode anode

FIG.6.4. (a) Schematic of a scintillation detector showing a scintillation crystal optically

coupled to a photomultiplier tube (PMT); (b)schematic of a microchannel plate-photomultiplier

PMTs are available with a large variety of specifications, including circular,

square or hexagonal photocathodes. Cathode diameters are in the range of ~10 to
~150mm. A ~50mm diameter PMT has a length of ~150mm including contact
pins. Pixelated multi-anode PMTs exist as well. To optimize time resolution,
special tubes are made with almost equal electron transit times to the anode,
independent of the cathode position where an electron is emitted. Although the
electron transit time is of the order of 30 ns, the transit time spread standard
deviation is not more than ~250 ps, and the signal rise time ~1.5 ns.
A PMT aimed at ultra-fast timing is the microchannel plate (MCP)-PMT.
For electron multiplication, it employs an MCP structure instead of a dynode
configuration (seeFig.6.4(b)). An MCP (thickness: ~1mm) consists of a large
number of closely packed hollow glass tubes (channel diameter: 550 m).
The inner surface of the tubes is covered with a secondary emission material,
viz. PbO. The glass surfaces on the front and back side are covered with metal
contacts. The MCP is placed in a vacuum, and a voltage of ~1000V is applied
between the contacts, positive on the back side. An electron that enters a glass
tube on the front side will hit the wall and secondary emission will occur. The
secondary electrons will be pulled to the back side by the electric field, hit the
channel wall and produce secondaries, etc. Eventually, they will leave the tube
at the back. Electron multiplication of ~104 can be obtained. In an MCP-PMT,
two MCPs are used at a close distance. The glass tubes are at an angle, thus
preventing ions from gaining too much energy. This structure of two MCPs is
called a chevron. At voltages of ~3000 V, stable gains of the order of 106 are
obtained. The advantage of the MCP-PMT is the short path length of the
electrons, resulting in transit times of a few nanoseconds and transit time spreads
of ~100 ps. MCP-PMTs are commercially available with circular (~10mm
diameter) and square photocathodes, the latter with multi-anode structures. The
sensitivities range from 115 nm (MgF2 window) to infrared.


Quantum efficiency (%)



Wavelength (nm)

FIG.6.5. Quantum efficiency as a function of scintillation wavelength for a blue sensitive

photomultiplier tube (full line), a photomultiplier tube with sensitivity extended into the
ultraviolet (dashed extension) and a silicon photodiode. Silicon based photon sensors

Although PMTs are used on a large scale in nuclear medicine, the relatively
large size, high voltages, small quantum efficiency and sensitivity to magnetic
fields are a reason to prefer the use of silicon photodiodes in some applications.
These diodes are usually of the pin structure (PIN diodes). They have a
thickness of ~2mm including packaging, and are circular, rectangular or square,
up to ~30mm30mm. Bias voltages are <150 V. The quantum efficiency
of silicon diodes can be >80% at longer wavelengths (Fig.6.5). The large
capacitance of 20300 pF, and leakage current, ~110 nA, are a disadvantage,
resulting in a significant noise level that negatively affects energy resolution in
An avalanche photodiode (APD) is the semiconductor analogue to the
proportional counter. A high electric field is created in a small zone where a
drifting electron can gain enough energy to produce an electronhole (eh)
pair. An avalanche will result. The critical field for multiplication is ~107 V/m.
The higher the voltage, the higher is the gain. Depending on the type, voltages
are applied in the range of 501500 V. Gains are in the range of M < 200 to
~1000. The gain lifts the signal well above the noise as compared with the silicon
diode. At a certain gain, the advantage is optimal. At very high electric fields,
spontaneous charge multiplication will occur. The corresponding voltage is


called the break-down voltage Vbr. For gains of M 105 106, an APD can be
used at voltages >Vbr, where it operates in Geiger mode. The pulses are equal in
magnitude. Signal quenching techniques have to be used. Circular and square
APDs are available with areas in the sub-square millimetre to ~1cm2 range.
Various pixelated APDs are available, e.g. of 4 pixels8 pixels at a pitch of
~2.5mm and a fill factor 40%.
In a hybrid photomultiplier tube (HPMT), the photoelectrons are accelerated
in an electric field resulting from a voltage difference of ~10kV, applied between
the photocathode and a silicon diode which is placed inside the vacuum enclosure.
The diode is relatively small, thus reducing the capacitance and, consequently,
the noise level. As the production of 1 eh pair will cost 3.6 eV, ~3000 eh pairs
are produced in the diode per impinging electron. Consequently, the signals from
one or more photons can be observed well separated. Equipped with an APD, an
overall gain of ~105 is possible. HPMTs have been made with pixelated diodes.
Window diameters are up to ~70mm.
The silicon photomultiplier (SiPM) is an array of tiny APDs that operate
in Geiger mode. The dimensions are in the range of ~20 m 20 m to
100 m 100 m. Consequently, the number of APDs per square millimetre
can vary from 2500 to 100. The fill factor varies from <30% for the smallest
dimensions to ~80% for the largest. The signals of all of the APDs are summed.
With gains of M 105106, the signal from a single photon can be easily observed.
By setting a threshold above the one electron response, spontaneous Geiger
pulses can be eliminated. The time spread of SiPM signals is very small, <100 ps.
Excellent time resolutions have been reported. Arrays of 2 pixels2 pixels
and 4 pixels4 pixels of 3mm3mm, each at a pitch of 4mm, have been
commercially produced. A 16 pixel16 pixel array of 50mm50mm has
recently been introduced. Blue sensitive SiPMs have a photon detection
efficiency of ~25% at 400 nm, including a 60% fill factor.
6.4.3. Scintillator materials Inorganic scintillators
In an inorganic scintillator, the bandgap has to be relatively large to avoid
thermal excitation and to allow scintillation photons to travel in the material
without absorption (Egap 4 eV). Consequently, inorganic scintillators are based
on ionic-crystal materials. Three steps for the production of scintillation photons
are considered (Fig.6.6): (i) interaction of radiation with the bulk material
and thermalization of the resulting electrons and holes on the energy scale,
electrons end up at the bottom of the conduction band and holes at the top of
the valence band; (ii) transport of these charge carriers to intrinsic or dopant


luminescence centres; (iii) interaction with these centres, i.e. excitation, relaxation
and scintillation. Using this model, the number of photons Nph produced under
absorption of a ray with energy E is:
N ph =

E gap

SQ (6.6)

The first term on the right is the number of eh pairs at the bandgap edge.
Typically, 2.5. S and Q are the efficiencies of steps (ii) and (iii).

Conduction band




Valence band


FIG.6.6. Energy diagram showing the main process steps in an inorganic scintillator.

For the most relevant scintillators, the wavelength at emission

maximum max, light yield Nph, best reported energy resolution at 662keV R662
and the decay time of the scintillation pulse are presented in the last four
columns of Table6.3. In the first columns, some material properties are given,
namely density, effective atomic number for photoelectric effect Zeff, attenuation
length at 511keV, 1/511, and the percentage of interaction by the photoelectric
effect at 511keV. These scintillators are commercially available. If hygroscopic,
they are canned with reflective material (Fig.6.4). Only BaF2 and BGO have
an intrinsic luminescence centre. The other scintillators have Tl+ or Ce3+ ions as
dopant luminescence centre. The cerium doped scintillators show a relatively fast
response of the order of tens of nanoseconds due to the allowed 5d 4f dipole
transition of the Ce ion. The transitions of the Tl doped scintillators are forbidden
and are, consequently, much slower. In general, mixed or co-doped crystals have
advantages for crystal growing, response time, light yield or afterglow effects.
Large variation is observed for light yields. This is mainly due to S < 1, i.e. there
are traps of different kinds, resulting in loss of eh pairs by non-radiative



transitions. Using Eq.(6.6) and the proper values of Egap, only LaBr3:Ce appears
to have S Q 1.

Zeff 1/511 Photoelectric max

(mm) effect (%) (nm) (photons/MeV) (%)

















4.3 800, 104









































































Hygroscopic. Organic scintillators crystals, plastics and liquids

The scintillation mechanism of organic scintillators is based on molecular
transitions. These are hardly affected by the physical state of the material.
There are pure organic scintillator crystals such as anthracene, plastics such
as polystyrene, and liquids such as xylene. Furthermore, there are solutions
of organic scintillators in organic solid (plastic) and liquid solvents. Typical
combinations are p-terphenyl in polysterene (plastic) and p-terphenyl in toluene.
There are also systems with POPOP (para-phenylene-phenyloxazole) added for
wavelength shifting. In general, organic scintillators luminesce at ~420 nm, have



a light yield of ~10 000 photons/MeV of absorbed ray energy and the decay
times are about 2 ns. The scintillators are usually specified by a commercial code. Storage phosphors thermoluminescence and
optically stimulated luminescence
A storage phosphor is a material analogous to an inorganic scintillator.
The difference is that a significant part of the interaction energy is stored
in long-living traps. These are the memory bits of a storage phosphor. The
lifetime must be long enough for the application considered. Readout is done
either by thermal stimulation (heating) or by optical stimulation. An electron is
lifted from the trap into the conduction band and transported to a luminescence
centre. The intensity of the luminescence is recorded. These processes have been
coined thermoluminescence and optically or photon stimulated luminescence.
Storage phosphors have been used for dosimetry for more than fifty years
(thermoluminescence dosimeter). In particular, LiF:Mg,Ti (commercial name
TLD-100) is widely used. The sensitivity is in the range of ~50 Gy to ~1 Gy. A
newer and more sensitive material is LiF:Mg,Cu,P (GR-200), with a sensitivity in
the 0.2 Gy to 1 Gy range. Recently, an optically stimulated luminescent material
has been introduced, Al2O3:C. The sensitivity is in the range of 0.3 Gy to 30 Gy.
Storage phosphors are also used in radiography.
APPLICATIONS, SCINT 2007, IEEE Trans. Nucl. Sci. 55 (2008) 10291564.
SCINT 2009, IEEE Trans. Nucl. Sci. 57 (2010) 11571520.
X- AND GAMMA-RAY DETECTORS (15th workshop), IEEE Trans. Nucl. Sci. 54 (2007)
(16th workshop), IEEE Trans. Nucl. Sci. 56 (2009) 16971884.
KNOLL, G.F., Radiation Detection and Measurement, 4th edn, John Wiley & Sons, New York
LEO, W.R., Techniques for Nuclear and Particle Physics Experiments, 2nd edn, Springer,
Berlin (1994).
CONFERENCE (annually), IEEE Trans. Nucl. Sci. (recent volumes).



RODNYI, P.A., Physical Processes in Inorganic Scintillators, CRC Press, Boca Raton, FL
SCHLESINGER, T.E., JAMES, R.B. (Eds), Semiconductors for Room Temperature Nuclear
Detector Applications, Academic Press, San Diego, CA (1995).
TAVERNIER, S., GEKTIN, A., GRINYOV, B., MOSES, W.M. (Eds), Radiation Detectors for
Medical Applications, Springer, Dordrecht, Netherlands (2006).


Joint Department of Physics,
Royal Marsden Hospital
and Institute of Cancer Research,
Rutherford Appleton Laboratory,
United Kingdom
Nuclear medicine imaging is generally based on the detection of X rays and
rays emitted by radionuclides injected into a patient. In the previous chapter,
the methods used to detect these photons were described, based most commonly
on a scintillation counter although there are imaging devices that use either gas
filled ionization detectors or semiconductors.
Whatever device is used, nuclear medicine images are produced from a
very limited number of photons, due mainly to the level of radioactivity that can
be safely injected into a patient. Hence, nuclear medicine images are usually
made from many orders of magnitude fewer photons than X ray computed
tomography (CT) images, for example. However, as the information produced
is essentially functional in nature compared to the anatomical detail of CT, the
apparently poorer image quality is overcome by the nature of the information
The low levels of photons detected in nuclear medicine means that
photon counting can be performed. Here each photon is detected and analysed
individually, which is especially valuable, for example, in enabling scattered
photons to be rejected. This is in contrast to X ray imaging where images are
produced by integrating the flux entering the detectors. Photon counting,
however, places a heavy burden on the electronics used for nuclear medicine
imaging in terms of electronic noise and stability.


This chapter will discuss how the signals produced in the primary photon
detection process can be converted into pulses providing spatial, energy and
timing information, and how this information is used to produce both qualitative
and quantitative images.
As described in Chapter6, the methods used for the detection of X ray and
ray photons fall into three categories, namely the scintillation counter, gas filled
detectors and semiconductors. Each of these techniques provides several detector
types and requires different electronics to produce and utilize the signals.
7.2.1. Scintillation counters
Figure7.1 shows a block diagram of a scintillation counter using a phosphor
and photomultiplier combination, together with the basic electronics required
to produce analogue and digital signals used to create an image. Table6.3
shows that the phosphors used in nuclear medicine can produce 150067000
optical photons per megaelectronvolt of energy deposited in the crystal and the
light emission time can vary from less than 1 ns up to ~1 s. Additionally, the
amplification of the optical signal by a photomultiplier can vary by an order of
magnitude or more depending on the photocathode quantum efficiency and the
number of dynodes. From this, it can be seen that the pulses produced by the
scintillation counter can vary substantially in both shape and amplitude, and that
the electronic devices used to manipulate these signals must be flexible enough
to account for these variations.
base and

analogue to

FIG.7.1. Block diagram of a scintillation counter and associated electronics.

If the signal from the photomultiplier tube (PMT) anode is small, a

preamplifier is needed prior to full amplification. This form of amplifier is usually
incorporated into the PMT electronic base tominimize the noise generated prior
to preamplification. Similar arguments apply to the use of solid state based light
sensors such as photodiodes when coupled to phosphors.


Both PMTs and photodiodes require voltage supplies to produce signals

in the case of a PMT, this voltage supply can be 13kV as each successive
dynode typically requires 100200V to produce sufficient amplification of the
electron signal. For a photodiode, the voltage required to totally deplete the
device is usually a few tens of volts for a simple photodiode and more for an
avalanche photodiode (APD).
7.2.2. Gas filled detection systems
Gas filled imaging systems convert the energy deposited by a ray photon
directly into ion pairs. It takes 2535 eV to produce a single ion pair, so the
primary signal from 99mTc will be 40005000 electrons. This signal will be
amplified in the gas detector using a high voltage (a few kilovolts) to produce
an electron avalanche of typically 106107 in a multiwire proportional chamber
(MWPC) (Fig.7.2). Typical dimensions of these devices for medical imaging are
between 30 and 100cm laterally by 1020cm in the direction of the ray.
These signals will clearly also need amplification if they are to be used as
analogue and digital output pulses for image formation.

Incoming ray
Cathode wire plane


Gas filled chamber

Anode wire plane

FIG.7.2. Schematic of a two-plane multiwire proportional chamber detecting a ray.

7.2.3. Semiconductor detectors

A simple diode can be thought of as a solid state ionization chamber where
the region between the pn junction acts as a reservoir of electronhole (eh)
pairs. Incoming radiation produces eh pairs in the diode, the number of which
is proportional to the energy deposited in the diode, and an array of diodes can
function as a radiation imaging device. The energy needed to produce an eh pair
is ~35 eV (seeTable6.2) and so the size of the pulses produced varies less than


for a scintillation counter. However, the signals will still require some form of
amplification to produce useful analogue or digital information.
Having briefly discussed the production of signals by the three major
ionizing radiation detection processes, it is necessary to understand how these
methods are used to produce images in nuclear medicine. The two main imaging
devices used are the gamma camera and the positron camera. For completeness,
autoradiography imaging of tissue samples containing radiotracers is also
Generally, both gamma camera and positron camera systems use
scintillation counters as the primary radiation detector because the stopping
power for X rays and rays is good in the high density scintillating crystals
used. However, there have been some examples of cameras using MWPCs and
semiconductors, and a brief description is provided here.
7.3.1. The gamma camera
Invented by Hal Anger, the gamma camera is usually based on the use of
a single large area (e.g. 50cm40cm of NaI(Tl)) phosphor coupled to up to a
hundred PMTs. The camera (Fig.7.3) can detect rays emitted by a radiotracer
distributed in the body. The lead collimator placed in front of the scintillation
counter selects the direction of the rays entering the device and allows an image
of the biodistribution of the tracer to be made.

Signal processing electronics

Incoming rays
FIG.7.3. Schematic of a gamma camera for detecting single photons.



The PMTs produce a signal that is proportional to the light generated in

the crystal. Positional information can be obtained by comparing the size of the
signals from different PMTs, whereas the energy information is related to the sum
of the PMT signals. The accuracy of both the energy and the positional information
depends on the stability of the signal production and also on the electronics used.
The amplitude of analogue signals from the PMTs depends on both PMT and
electronic stability, and the noise generated in the signal amplification process. The
noise is kept as low as possible by amplification close to each PMT. The amplified/
summed signal from the photomultiplier can be converted to a digital pulse train
using an analogue to digital converter (ADC), where the number of pulses is
proportional to the pulse height or charge in the signal. This signal is used to select
events in which most of the energy of the ray is detected by the camera, a useful
tool to reduce the effect on image production of rays scattered in the body prior
to detection. It is important at this stage to include only those signals from PMTs
that provide information above the intrinsic noise level of the electronics this is
done using some form of signal thresholding, such as a comparator.
Traditionally, the individual PMT pulses are digitized close to the PMTs
and these signals are analysed using capacitor or resistor circuits to determine
the positional information that is then sent to the computer system to form the
image. The centroid of the energy/pulse height information provides a position
that is closely related to the point at which the ray enters the crystal. Recent
improvements for calculating this position based on the digital outputs from the
PMTs uses nearest neighbour calculations based on a stored reference map of
positional information. These methods provide more accurate estimates of the
incident radiation entry point but are more demanding computationally.
Thus, the accuracy of the image production process is very much determined
by the initial signal sizes and the subsequent amplification and digitization.
7.3.2. The positron camera
The positron camera is used to simultaneously detect the two annihilation
photons produced by positron emitting tracers distributed in the body. The
detectors are usually made of many thousands of small scintillating crystals
coupled to up to a thousand PMTs. This means that there are many more amplifiers
which may have to function at higher count rates than those used in a gamma
camera. The detection of these two rays requires the addition of coincidence
circuits used to select the pulses from a single annihilation event. Figure7.4
illustrates the format of a positron camera based on multiple scintillating counters
in which the signals can be read out to form a 3D image.
The main difference in the electronics between the positron camera and
the gamma camera is the large number of PMT channels involved and the count


rates achieved in both cases, factors of 1020 are not unusual. In addition, the
pulses from a positron camera must be carefully shaped to allow accurate timing
information to be made in coincidence circuits to ensure that both annihilation
photons from a single annihilation event are detected. Time jitter in the pulses
will affect this process, allowing random photons from multiple nucleic decays
to be included in the data acquisition. In addition, the recent introduction of
so-called time of flight cameras requires very accurate (sub-nanosecond) timing
to be made between the two annihilation photons.



Energy 1

Digital 1




Digital 2
Energy 2

FIG.7.4. Typical signal processing in a phosphor/photomultiplier tube based positron camera

pulses from two opposing detectors in the camera array are amplified, digitized, checked for
coincidence and then used to provide positional and energy information for the event detected.

7.3.3. Multiwire proportional chamber based X ray and ray imagers

As shown in Fig. 7.2, an X or photon generates an ionization signal in
the gas that is detected by the anode and cathode wire planes. The high voltages
across the wire planes cause electron avalanches close to the nearest wires and
these signals can be detected and amplified at either end of the wire.
In the PETRRA positron camera (Fig. 7.5), the initial ray detection is
performed using blocks of barium fluoride. The vacuum ultraviolet produced
in the crystal photo-ionizes the gas producing electrons that are subsequently
amplified in the gas by a series of wire planes. The MWPC positional
information is read out using delay lines coupled to the cathode wires. Signals
are induced in the delay lines and detected at either end using amplifiers that
produce signals with low time jitter. These signals are passed to constant fraction
discriminators (CFDs) to produce fast timing signals and then to time digitizers.
The time difference between the arrival of the signals at the two ends of the


delay line is measured by the time to digital converters (TDCs) and provides
the positional information the accuracy of this information depends on the
intrinsic properties of the delay lines and the spread of the signal at the wire
planes, and in this system is ~4mm. Pulses produced after the gas amplification
region are used to provide the fast coincidence trigger to read the data into the
computer a timing resolution of ~23 ns is readily achievable.

From second detector


Data enable



Pulse height

B Gamp



Barium fluoride crystal

Gas amplification wire planes
Gate wire plane
Cathode wire plane
Anode wire plane
Coincidence box
Constant fraction discriminator
Time digitizer

FIG.7.5. Schematic of the pulse production and readout system for the PETRRA positron
camera wire planes are shown as dotted lines.

A further process tominimize the recording of single ray events in this

system is the use of an electronic gate in which the transport of ionization
signals is controlled by an additional wire plane. This allows passage of the
ionization to the anode/cathode part of the chamber only if two MWPC detectors
have been triggered in fast coincidence. Anode signals can be used to measure
the pulse height produced by each detected ray. Overall, the electronics of this
camera has to manage count rates of several megahertz.
Similar detectors have been developed for imaging low energy rays,
animal imaging and tissue autoradiography.
7.3.4. Semiconductor imagers
There have been several attempts to make nuclear medicine imaging devices
using semiconductors as a ray camera. The need for a high Z material means
that presently only germanium and cadmium zinc telluride (CZT) have potential
as the primary ray detector. Germanium (in the form of GeLi) has been used as
a 2D strip detector where the signals from amplifiers at the end resistor chains


could be used to determine the position of any interaction in the sensor. However,
the modest stopping power of the material coupled with the need for a cryostat to
reduce the intrinsic noise of the detector made this design impractical.
More practical systems based on room temperature operation of CZT
have been developed by GE (the Alcyone system) and Spectrum Dynamics (the
DSPECT system). In the case of the latter system designed specifically for cardiac
imaging, ~1000 individual small CZT crystals are coupled to a tungsten collimator
providing an intrinsic spatial resolution of 3.54.2mm full width at half maximum
and a sensitivity of approximately eight times that of a scintillator based camera
most of the increases in sensitivity are due to the collimator design.
Silicon photodiodes have been used as an alternative to PMTs for both
gamma camera and positron camera designs. Here, APDs have been coupled to
phosphors and because of their small size, a truly digital camera design is possible.
In practice, due to the cost of APDs, only small systems have been developed. The
recent development of silicon photomultipliers promises further improvements in
nuclear medicine imaging.
7.3.5. The autoradiography imager
Autoradiography is based on the use of radioactive labels to determine the
microscopic distribution of pharmaceuticals in tissues excised from humans
or animals. A major use in humans is to detect areas of malignancy or tissue
malfunction. In animals, the method is used to track the uptake of drugs, for instance.
The pharmaceuticals are usually labelled with long lived radiotracers that have a
short range emission or low energy X ray or ray emission. Typical examples of
tracers used are 3H, 14C, 32P, 33P and 125I. Autoradiography imagers are required to
detect the emissions with high efficiency as the levels of uptake in tissue samples
are often very low. The gold standard for tissue radiography is film emulsion which
produces a high resolution (m) image of tissues, although these detectors have
low efficiency for detecting the radiation involved. Images can take days to weeks
to produce and this can be a severe limitation if diagnostic information is desired.
Digital autoradiography systems based on the use of thin phosphors, gas filled
detectors and silicon wafers can be 50100 times more efficient although the spatial
resolution is limited to typically a few tens of micrometres.
A phosphor based imager may use a very thin (50100 m) material such as
GADOX or CsI(Tl) coupled to a high resolution sensor, such as a microchannel
plate, a charge coupled device or a complementary metal oxide semiconductor
APD (Fig.7.6).




igure 7.4

igure 7.6

igure 7.7

igure 7.8

Energy 2


Beta particle
Segmented CsI(Tl)
Read out

Read out

FIG.7.6. Beta particle detection in an autoradiography system based on a segmented CsI(Tl)

phosphor coupled to a complementary metal oxide semiconductor avalanche photodiode

The limitations of these devices are mostly the pixel size of the sensor
and the noise in the sensor. Amplifiers with low noise are required and room
temperature operation desirable. Such a device can have a resolution of <50 m.

MWPC based autoradiography imagers have been built in which the

sample is placed in intimate contact with the chamber gas such a device will
Ihave a resolution of a few hundred micrometres.
More recently, direct detection
of particles and X rays has been performed using charge coupled devices and
+ complementary metal oxide semiconductor APDs. The advantage of such devices
is that the spatial resolution can be improved further (down to a few micrometres)
and 3H can be imaged.

As discussed
above, the primary signals from the radiation detectors are
generally small and need to be amplified without the injection of high levels
of noise into the signal readout system.
A preamplifier is needed prior to the
main amplification process if the signals
from the detector are very small, for
example, when aRPMT has insufficient dynodes to provide a large output pulse.
Preamplifiers are usually mounted immediately next to or as part of the output
stage of the detector tominimize the noise produced prior to full amplification.
The main amplifier can then be used to maximize and shape the signal (via
current and/or voltage gain) without over-amplifying noise.
7.4.1. Typical amplifier
The output current from a PMT is directly proportional to the amount
of light received from the phosphor. Although the PMT amplifies the electron
signal produced at the photocathode by a large factor, the current produced at the
anode is still very small. Amplifiers for PMTs are specially designed to transform
this current into voltage which can be directly input into an analogue to digital





converter or a comparitor. In order to achieve the optimum signal to noise ratio,

the output current pulse is integrated in a capacitor, the resulting voltage Coincidence
circuit of
the output signal. The capacitor is normally arranged in the feedback circuit
a wide bandwidth voltage amplifier chosen to have high input impedance and
ADC(tens of kiloherz
an extremely small input current. As data rates can be high
to megaherz), the operational frequency range of the amplifier must be able
to cope with these rates. Ideally, the PMT anode is connected directly to the
D2 a high Amplifier
charge amplifier input, with
value resistor providing a DC return path.
Capacitative coupling can be a problem at low frequencies where the signal may
7.4 The charge amplifier integrates the current from the PMT, producing
an output voltage pulse. Figure7.7 illustrates a typical charge amplifier where the
output voltage Vout is given by:
Vout =

Beta particle

I (t ) d t =

The configuration has negative feedback that increases the effective input
Read out
capacitance by a factor equal to the gain of the amplifier. This ensures that almost
all of the current flows into the amplifier even though the PMT and wiring can
7.6 capacitance. The feedback also reduces the output impedance, so
that the amplifier acts as a voltage source.



FIG.7.7. A directly coupled charge amplifier producing a voltage output by integrating the
Figure 7.7
current produced in the photomultiplier tube.

The shape of the output pulse is important for the measurement of both
analogue and digital information,
C and is defined by the output stage of the
amplifier (Fig.7.8). The amplified signal is first passed through a CR (high pass)
R the low frequencies,
filter which improves the signal to noise ratio by attenuating

Figure 7.8



Read ou




Figure 7.7
which contain a lot of noise and very little signal. The decay time of the pulse is
also shortened by this filter.
Figure 7.8

FIG7.8. A CRRC pulse shaping circuit.

Before the output of the amplifier, the pulse passes through an RC (low pass)
filter which improves the signal to noise ratio by attenuating high frequencies,
which contain excessive noise. The pulse rise time is lengthened by this filter.
The combined effect produces a unipolar output pulse and with suitably chosen
values, has an optimal signal to noise ratio.
7.4.2. Properties of amplifiers
The most important properties of an amplifier are gain, bandwidth, linearity,
dynamic range, slew rate, rise time, ringing, overshoot, stability and noise:
The gain of an amplifier is defined as the log ratio of the output power/
voltage Pout to the input power/voltage Pin and is usually measured in
gain [dB]=10log(Pout/Pin) (7.2)
Gain in charge amplifiers is often expressed in millivolts output per
picocoulomb input charge.
The bandwidth of an amplifier is defined as the range of frequencies that
the amplifier operates and is often determined by frequencies at which the
power output drops to half its normal value (the 3 dB point). This is an
important feature of an amplifier attached to a high count rate detector as
required in positron emission tomography (PET) imaging, for example.
Amplifier linearity is limited when the gain of the amplifier is increased
to saturation point, resulting in output pulse distortion. Clearly, this is
important if the dynamic range of the pulses produced by the detector is
large. Dynamic range is defined as the ratio of the smallest and largest


useful output signals, with the former limited by the noise in the system and
the latter by amplifier distortion.
Rise time is often defined as the time taken for the output pulse to increase
from 1090% of its maximum and is a measure of the speed or frequency
response of the amplifier.
Slew rate is the maximum rate of change of the shape of the output
signal for the whole range of input signals, usually expressed in volts per
microsecond. This is very important if timing information is needed from
the detector as a poor slew rate will distort the bigger signals, making
them unsuitable for fast timing, as in PET. For PET applications, amplifier
rise times of the order of a few nanoseconds are needed, with no shape
distortion resulting from slew rate even on the biggest pulses.
Ringing is a problem when an amplifier produces a pulse that either
oscillates before reaching its maximum value or where the tail oscillates
before reaching the baseline. This can be a serious problem if timing
information is required or if the oscillations produce multiple triggers of the
output electronics downstream of the amplifier.
Stability is clearly an important parameter for an amplifier if the output
signals are to be used for either analogue or digital purposes. It is essential
that the amplifier output does not vary significantly for a given input
signal as the processes used to determine positional, energy and timing
information rely on the output for a given input being constant both in
offset, amplitude and shape. Factors that affect stability are numerous but
prime examples are variations in temperature, supply voltage and count rate
as well as long term drift.
Noise is a major impediment to the production of images using any of the
devices discussed above. Examples include thermal noise caused by the
thermal movement of charge carriers in resistors, shot noise caused by a
random variation in the number of charge carriers and flicker or 1/f noise
caused by the trapping or collisions of charge carriers in the structure of
the silicon used in the electronics. These sources combine to produce a
variation in the output signal of the combined detector/electronics system
that can affect the quality of images produced. The root mean square noise
of a system is defined as the square root of the absolute value of the sum of
the squares of the noise variances.
For a system using PMTs, the dominant noise component is that associated
with the number of photoelectrons produced at the photocathode as this is
amplified by the gain of the PMT dynode chain and subsequent electronics. For
a gas filled detector, the equivalent is the number of primary electrons produced



at the first stage of the ionization process and for a silicon detector the important
parameter is the initial number of eh pairs produced.
Once an amplified signal has been produced, it is then used to generate
both analogue and digital information about the detected event. The analogue
signal will relate to the energy deposited in the detector and is used, for example,
tominimize the number of scattered rays accepted into the image production
process. The digital signal is used to produce spatial and timing information.
7.5.1. Analogue signal utilization
The analogue information is generated by sending the pulse from the
amplifier into a single or multichannel pulse height analyser. In a gamma camera,
several energy windows are available, whereby the pulse height or charge is
compared with preset values that correspond to the known energies of the rays
being detected. In the simplest case for imaging a single energy ray emission,
two thresholds can be set to reject pulses that are above or below these values

Threshold 1
Threshold 2
Input pulse

Output pulse

FIG.7.9. A single channel pulse height analyser an output pulse is produced when the input
pulse is between the two thresholds. This system also functions as a single channel analogue
to digital converter.

When a radiotracer that emits several different energy rays is being used,
multiple thresholds can sort the information into several channels or images.
7.5.2. Signal digitization
Analogue signals are converted into digital signals that are subsequently
used to provide spatial and temporal information about each detected event.







Digital pulse train to computer

FIG.7.10. Ramp-based single slope converter system for digitizing analogue pulses.

This is done using an ADC. The simplest method of digitizing an analogue

signal is by using a single slope converter (Fig.7.10). For this, a ramp signal is
generated and at the same time a clock producing digital output pulses is started.
When the ramp signal exceeds the input pulse, the clock is stopped and the

FIG.7.11. Pulse sequence from the system illustrated in Fig.7.10.

number of pulses generated corresponds to the amplitude of the signal. The faster
the clock, the higher the accuracy of the digitization achieved. This is a relatively
simple and low cost solution but is slow as the time taken to digitize the pulse is
2N clock cycles. The pulse sequence producing the digital output is shown


in Fig.7.11. An important feature is that the analogue pulse shape must be

constant to allow accurate conversion. It is clearly possible to have more than one
ramp signal to provide several digitization regions if greater or lesser accuracy is
needed in any region.
A faster method of analogue to digital conversion is possible by using a
flash ADC. This is done using a large number of comparators (seeFig.7.12),
each having a different reference level. The output from each is the input into a
logic box that produces the multiple bits of the digital signal. If an N bit output
is needed, then 2N 1 comparators are needed. The method is fast as conversion
takes a single clock cycle but the system is complex and expensive and consumes
a lot of power. Typically, between 8 and 12 bits may be needed for nuclear
medicine imaging.

1st bit
2nd bit

3rd bit

Nth bit
Reference level 2N 1
FIG.7.12. Schematic of a FLASH analogue to digital converter producing N bits of digital

7.5.3. Production and use of timing information

In PET systems, the timing of events is very important as only pairs of
annihilation photons from the same radioactive decay contribute positively to
the image. As the single count rates in a PET scanner may be very high, fast
timing is required for coincidence imaging and time of flight measurement.
Coincidence timing systems can be based on the timing taken from the front edge
of two pulses, from a zero crossing point of the differentiated pulses or by using
a constant fraction method. The main problem with using a simple front edge
trigger is that the variation in pulse height of the analogue signals produces a
large variation in timing. Figure7.13 shows how the two pulses from detectors in
a PET system are used to generate a coincidence with CFDs.


Input 1

Input 2

FIG.7.13. The use of constant fraction discriminators (CFDs) to generate a fast coincidence
output the insert shows how the trigger point is set by a constant fraction of the pulse height.

The trigger points for the timing occur at a constant fraction of the shaped
analogue signal, so that the timing is not affected by the different signal pulse
heights. In this example, CFD1 generates a gate with a width set to more than
twice the measured timing resolution of the detectors. If the pulse from CFD2
falls within this gate, a coincidence (AND) output is generated; otherwise, the
event is rejected.
An alternative method of determining the timing from a pulse is to use the
zero crossing technique (Fig.7.14). In this method, the pulse is differentiated to
produce a bipolar pulse the timing is taken from the point where the pulse
crosses a reference line that is usually tied to ground hence, the zero crossing.
Again it is important that the pulse shapes are carefully controlled tominimize
jitter in the timing information.

FIG.7.14. A timing signal generated from the zero crossing point of a differentiated signal.

If the timing information is to be stored, then the two pulses from the
CFDs can be input into a TDC. In this case, the first pulse starts and the second
one stops a clock the number of pulses generated is proportional to the time


difference between the pulses. If a fast clock is used, excellent timing information
is available for use in time of flight calculations, for example.
7.6.1. Power supplies
Low voltage supplies are used to provide the power input for semiconductor
systems where a few tens of volts are sufficient. In some cases, batteries may
provide enough power but the need to maintain a constant current and voltage
makes this a modest solution. Usually, a low voltage supply converts mains AC
power, typically 240V (or 110 V), into DC voltages of, for example, 15V
and 5V to provide the line voltages for transistors and diodes. This is done by
combining a transformer, which reduces the voltage, and a rectifier, typically a
diode which allows only one half of the AC signal to pass this is half-wave
rectification (Fig.7.15).
Full-wave rectification is achieved by using a diode bridge that allows both
halves of the AC signal to be used, with one half being inverted. The oscillations
are removed using a filter, usually capacitors. The smoothest DC output is
provided by using a three phase AC input. The output is usually passed through
a voltage regulator to stabilize the voltage and remove the last traces of ripple.
Half-wave rectification

3-phase full-wave rectification

AC input

Full-wave rectification
FIG.7.15. Conversion of AC into DC using a transformer and rectifier system.

PMTs and MWPCs require power supplies that can provide voltages up to
several kilovolts. For example, each pair of dynodes in a PMT usually has at least


100V between them and even more may be used between the photocathode and
the first dynode to maximize the early gain of the PMT. These power supplies
usually have an oscillator and step up transformer operating at high frequency
to provide the drive plus a voltage multiplier consisting of a stack of diodes and
7.6.2. Uninterruptible power supplies
This form of support is needed for an imaging system to overcome loss of
power during periods of mains supply interruption. In this case, the output power
comes from the storage battery via some form of inverter. While the mains power
is available, it charges the battery as well as providing power to the imaging
device. If the mains supply is interrupted, the battery continues to provide support
to ensure that the imaging system can continue to be used. The size of the battery
support system depends directly on how long backup is needed or how long, for
example, it takes the operator to save data and shut down the system. As in most
imaging environments, the mains is replaced by a generator supply. The period
of support is often short but usually several hours of supply is available from an
uninterruptible power supply.
7.6.3. Oscilloscopes
In order to optimize the use of pulse generating equipment, an oscilloscope
is essential. This type of device allows the pulses from the detectors to be
displayed at various stages of generation prior to their use in image production.
For example, the pulse sequences illustrated above can be displayed on an
oscilloscope and this allows the equipment to be adjusted to provide the optimum
analogue and digital pulse sequence, shape and size.
An oscilloscope allows the pulses to be displayed on a 2D display, usually
with the vertical axis representing voltage (pulse height) and the horizontal
axis time. In addition to the amplitude of the pulses, the oscilloscope display
can be used to analyse the frequency of the signals being studied and also to
detect any pulse distortion such as oscillation or saturation. In an advanced form,
the oscilloscope can function as a spectrum analyser over a wide range of pulse
The original oscilloscopes were based on a cathode ray tube to display
the pulses but more modern systems use liquid crystal displays connected to
ADCs and other signal processing electronics. To the user, the oscilloscope will
present as a box with a display screen, input connectors and various controls.
The input from equipment can be done either directly using connecting cables/
sockets or through probes, often into a high impedance (e.g. 1M) or, for high


frequency signals, 50 . The trace on the oscilloscope screen is adjusted by

various controls. The timebase control can adjust the horizontal display between,
for example, 10 ns up to seconds and the pulse height control from millivolts to
volts. Other controls include a beam finder, spot brightness and focus, graticule
control (to provide a visual measurement grid), pulse polarity and trigger level
controls, horizontal and vertical extent and position controls, selection of trigger
source (particularly useful for pulse coincidence display) and sweep controls to
provide single, multiple and delayed sweeps, for example. An example of an
oscilloscope that can be used for examining pulses from imaging equipment is
shown in Fig.7.16.

FIG.7.16. The front panel of the Tektronix 465 oscilloscope.

More recently, it has been possible to install oscilloscope software onto a

computer to provide a low cost solution for pulse display.
The information provided above gives a general overview of the equipment
and electronics used in nuclear medicine imaging. In some cases, manufacturers
have developed special multichannel electronics readout systems tailored to the
detectors. These systems include the individual electronics elements described
above in a compact design that increases speed and accuracy. Such systems are
usually specific to the device involved.



HOROWITZ, P., HILL, W., The Art of Electronics, Cambridge University Press (1982).
INIEWSKI, K. (Ed.), Medical Imaging: Principles, Detectors, and Electronics, John Wiley and
Sons, Hoboken, NJ (2009).
TURCHETTA, R., Electronics signal processing for medical imaging, Phys. Med. Imaging
Appl. 240 (2007) 273276.
WEBB, S., The Physics of Medical Imaging, Hilger (1988).


Department of Radiology,
University of Pennsylvania,
Philadelphia, Pennsylvania,
United States of America
8.1.1. Generic nuclear medicine imagers
The generic nuclear medicine imager, whether a gamma camera, single
photon emission computed tomography (SPECT) system or positron emission
tomography (PET) scanner, comprises several main components: a detection
system, a form of collimation to select rays at specific angles, electronics and a
computing system to create the map of the radiotracer distribution. This section
discusses these components in more detail.
The first stage of a generic nuclear medicine imager is the detection of the
rays emitted by the radionuclide. In the case of PET, the radiation of interest are
the 511keV annihilation photons that result from the interaction of the positron
emitted by the radionuclide with an electron in the tissue. For general nuclear
medicine and SPECT, there is one or sometimes more than one ray of interest,
with energies in the range of <100 to >400keV.
The rays are detected when they interact and deposit energy in the
crystal(s) of the imaging system. There are two main types of detector: crystals
that give off light that can be converted to an electrical signal when the ray
interacts (scintillators) and semiconductors, crystals that generate an electrical
signal directly when the ray deposits energy in the crystal. Scintillation detectors
include NaI(Tl), bismuth germanate (BGO) and lutetium oxyorthosilicate (LSO);
semiconductor detectors used in nuclear medicine imagers include cadmium zinc
telluride (CZT). Radiation detectors are described in more detail in Chapter6.
When a ray interacts in a scintillation crystal, it deposits some or all
of its energy. This energy is re-emitted in the form of light with a wavelength
dependent on the crystal material but not on the energy of the ray. The more
energy deposited in the crystal, the greater the intensity of the light emitted.
Scintillation crystals are coupled to photomultiplier tubes (PMTs), which serve


to convert the scintillation light into an electrical signal. If scintillation light

strikes the photocathode of the PMT, electrons are emitted from the photocathode
by the photoelectron effect. The number of photoelectrons emitted depends on
the intensity of the scintillation light and, therefore, the energy deposited in the
crystal. The energy required to produce a single photoelectron is ~1000 eV, so
only a few hundred to a thousand electrons are produced for each ray that
interacts, well below the number needed to produce a measurable current. The
PMT contains approximately ten stages that serve to increase the number of
electrons by secondary emission of electrons from these dynodes. The signal at
the output of the PMT is a measurable current, the amplitude of which is still
proportional to the energy deposited in the crystal.
Semiconductor detectors operate differently: the ray still deposits some
or all of its energy in the crystal through photoelectric absorption or, more
likely, Compton scattering interactions. However, that energy is not re-emitted
as scintillation light; instead, it creates electronhole (eh) pairs that are then
collected by application of an electric field to create a measurable signal. The
energy required to create an eh pair is ~3 eV, so many more charge carriers are
created in semiconductor detectors than in scintillators (seeChapter6).
While the exact implementations vary from system to system, the
electronics of nuclear medicine imagers have several common functions: they
determine the location of interaction of the ray in the detector, calculate the
energy deposited in the crystal and ascertain whether that energy falls within a
prescribed range of desirable energies, and for PET systems, measure the times
that the two annihilation photons interacted and evaluate whether the difference
in those times falls within a desired timing window to have both come from
the same annihilation event (i.e. from the same positron decay). If an event is
determined to be valid, its position (and sometimes the energy and timing
information) is sent to the computer to be stored along with the information for
the many other valid events.
In order to create an image of the distribution of radiotracer, the measured
locations of interaction of the rays must be converted to a 2D or 3D map
through image reconstruction. For 2D planar imaging with a stationary gamma
camera, this can be as simple as displaying the number of events at each detector
position. For PET or SPECT imaging, where measurements are made for many
views around the subject, the data must be combined through a reconstruction
algorithm. These techniques range from analytical methods such as filtered
back projection to iterative algorithms where estimates of the distribution are
calculated and refined based on a model of the imaging system. Not all events
accepted are actually useful events with accurate position, energy and timing
information. To obtain quantitative images (i.e. images whose counts are directly



related to the amount of activity at each location), corrections must be applied for
these unwanted events as part of the reconstruction process.
Performance measures aim to test one or more of the components, including
both hardware and software, of a nuclear medicine imager.
8.1.2. Intrinsic performance
There are two general classes of measurements of scanner performance:
intrinsic and extrinsic. Intrinsic measurements reflect the performance of a
sub-part of the imager under ideal conditions. For example, measurements
made on a gamma camera without a collimator will describe the best possible
performance of the detector without the degrading effects of a collimator,
although the collimator is essential for clinical imaging. For a PET scanner,
intrinsic performance is often determined for a pair of detectors, rather than
the entire system. Intrinsic measurements are useful because they reflect the
best possible performance and can help isolate the source of any performance
degradations observed clinically. However, these measures are typically
performed under non-clinical conditions and will not reflect the performance
of the nuclear medicine imager for patient studies. Intrinsic measures also tend
to be measurements of an isolated characteristic of the system, rather than its
impact on imaging studies. They reflect the limits of performance achievable by
the detection system and electronics without collimators or image reconstruction.
8.1.3. Extrinsic performance
Extrinsic, or system, performance measures are made on the complete
nuclear medicine imager under conditions that are more clinically realistic,
although even these measures may not show the full clinical performance of
the system. On a gamma camera, extrinsic measurements are made with the
collimator in place; for SPECT and PET systems, the performance is often
measured on the reconstructed image. The extrinsic performance of a system
gives an indication of how well all of the components of the imager work together
to yield the final image. As most extrinsic performance measurements attempt to
isolate a single aspect of imaging performance (e.g. spatial resolution, count rate
performance, sensitivity), the conditions of these measurements generally do not
match the conditions encountered in patient imaging studies. However, the results
of extrinsic performance measurements are generally good indicators of clinical
performance or may provide useful information about system optimization for
clinical studies.




8.2.1. Energy spectrum
The amplitude of the signal from the detector depends on the energy
deposited in the crystal. If the number of measured events with a given
amplitude is plotted as a function of the amplitude (Fig.8.1), the result is an
energy spectrum. The shape of the energy spectrum depends on the radiotracer
and rays emitted through its decay and the characteristics of the detector
material, but all energy spectra have common features. There is one (or more
than one) peak, called the photopeak, where the ray deposited all of its
energy in the detector through one or more interactions. There is also a broad,
lower energy region that reflects incomplete deposition of the rays energy
in the detector and/or Compton scattering of the rays in the body with the
subsequent loss of energy before detection. Even in the absence of scattering
material (i.e. for a point source in air), the photopeak is not a sharp peak but
is blurred. This broadening, which depends on the properties of the detector,
is due to statistical fluctuations in the detection of photons and conversion of
the energy deposited in the crystal into an electrical signal. This effect is larger
for scintillation detectors than for semiconductors. With scintillation detectors,
there are several steps in the conversion process that are subject to statistical
fluctuations, including the conversion of the rays energy into scintillation
light, collection of the scintillation light and conversion into photoelectrons at
the PMTs photocathode, and multiplication of those photoelectrons at each
dynode in the PMT. For semiconductor detectors, statistical uncertainty is
introduced in the number of eh pairs created when the ray deposits its energy
and in the collection of these pairs.
The goal of nuclear medicine imaging is to map the distribution of
radiotracers, so only rays that do not interact in the tissue before reaching the
detectors are useful; any rays that scatter in the body first change their direction
and do not provide an accurate measurement of the original radionuclides
location. Unscattered photons are those rays with energies in the photopeak.
Nuclear medicine imagers accept events whose energies lie in a window
around the photopeak energy in order to reduce the contribution of lower energy,
scattered rays. For PET scanners, a typical energy window is 440650 keV
for LSO detectors; for gamma cameras based on NaI(Tl) detectors, it is 15% of
the photopeak energy (e.g. 129.5150.5keV for 140keV rays from 99mTc, and
6882keV for the characteristic X rays from 201Tl with a 20% window).



Number of events


FIG.8.1. An example of an energy spectrum, defined as the number of measured events with
a given amplitude plotted as a function of the amplitude, where the amplitude depends directly
on the energy deposited in the crystal.

8.2.2. Intrinsic measurement energy resolution

The intrinsic ability of a detector to distinguish rays of different energies is
reflected in its energy resolution. The energy resolution of a detector is defined as
the full width of the photopeak at one half of its maximum amplitude, divided by
the energy of the photopeak, and is typically expressed as a percentage. A smaller
energy resolution value means that the detector is better able to distinguish
between two rays whose energies are close to each other. The energy resolution
depends on the energy of the ray approximately as ( + E)1/2/E and, therefore,
the energy of the ray source must be specified when quoting the energy
resolution of a system. The energy resolution worsens at lower energies because
fewer photoelectrons are detected (in scintillation detectors) or eh pairs are
created (for semiconductors), so the statistical fluctuations in the measured signal
are greater. In addition, the energy resolution of a complete imaging system is
typically worse than that of small individual detectors due to slight differences in
operating characteristics between detectors.


Only rays that have not scattered in the body will provide accurate
information about the radiotracer distribution. Accordingly, the energy window
is optimal if it includes as many photopeak events as possible, since they are
more likely not to have interacted with the tissue, and as few lower energy events
as possible, since they are more likely to be the result of one or more Compton
scatter interactions in the tissue. As the energy resolution worsens, however, it
is necessary to accept more low energy events because the photopeak includes
lower energy rays. For example, for detection of 511keV annihilation photons,
the lower energy threshold for BGO (1520% energy resolution) was typically
set to 350380keV, while that for LSO (12% energy resolution) is 440460keV
and for LaBr3 (67% energy resolution) the lower energy threshold can be set as
high as 480490keV without loss of unscattered rays.
8.2.3. Impact of energy resolution on extrinsic imager performance
The energy resolution is an intrinsic measure of detector performance; it
defines theminimum width of the energy window for a given radiotracer. The
energy window in turn affects the amount of scattered photons accepted. The ratio
of scattered events to total measured events, the scatter fraction, is an extrinsic
performance characteristic that is of concern, especially for quantitative imaging.
In PET systems, for example, the clinical image is assumed to be linearly related
to the activity uptake; because scatter adds a smoothly varying background to the
image, it degrades the quantitative accuracy of the image and adds to the image
noise, even when accurately estimated and subtracted.
There are two major types of scattered event, those where the initial ray
scattered in the body and those where the ray was not completely absorbed in
the detector but instead scattered, losing some but not all of its energy. In both
cases, the measured energy of the ray is lower than the energy of the original
photon because some energy is given up to the electron, and the measured
position may no longer be related to the original source of the ray because the
scattered photon does not travel along the same direction as the original ray. For
typical patient sizes, scattering in the body is much more significant than detector
The scatter fraction is an extrinsic performance measure that describes the
sensitivity of a nuclear medicine imager to scattered events. The measurement
involves imaging a line source in a uniformly filled phantom of a specified size
at a low activity level, where scattered and unscattered events can be reasonably
well differentiated. As the amount of scatter depends on the size and distribution
of scattering material in the scanner, the measured scatter fraction cannot be used
to infer the amount or distribution of scatter in patient images. However, it is a
good indicator of the relative sensitivity of the system to scatter.


The scatter fraction is directly related to the energy resolution of the

system in the sense that the energy resolution determines the energy window,
in particular the lower energy threshold. This determines the imagers ability to
exclude scattered events. However, good energy resolution does not lead to a low
scatter fraction unless the energy window used is made appropriately narrow; a
scanner with 7% energy resolution will accept approximately as much scatter
as one with 12% energy resolution if both systems have the same lower energy
threshold. For this reason, measurement of the scatter fraction is a more clinically
relevant parameter than the energy resolution.
8.3.1. Spatial resolution blurring
The spatial resolution of a nuclear medicine imager characterizes the
systems ability to resolve spatially separated sources of radioactivity. An
individual point source of activity does not appear at a single pixel in the image;
rather, it is blurred over several pixels, largely due to statistical fluctuations in the
detection of the rays. Sources whose measured activity distributions overlap
cannot be distinguished as distinct sources and instead appear as a single, broad,
low contrast source.
In addition to blurring small structures and edges, resolution losses also lead
to a decrease in the contrast measure in these structures and at boundaries of the
activity distribution. Activity in small structures is blurred into the background
and vice versa. Areas of increased or decreased uptake are less easily detected
because of this loss of contrast (the partial volume effect).
In imagers composed of many small crystals, the spatial resolution of
the system is limited by the size of the detector elements. In gamma cameras
with a single, large crystal coupled to an array of PMTs, the spatial sampling
of the crystal determines the best spatial resolution achievable. The smaller the
crystal element or the more finely sampled the detector, the better an event can be
localized and the better the spatial resolution will be.
For a given size and sampling, crystals of different materials will have
different spatial resolutions. This is because rays do not interact at the surface of
a crystal but penetrate the crystal before interacting. If a crystal has a low density
and low atomic number Z, rays will travel further before interacting, compared
with a high density, high Z material. The ability to stop rays is referred to as
the materials stopping power; detectors with higher stopping powers will have
more accurate spatial localization than those with low stopping power because
there is less inter-crystal scatter. The effect of stopping power becomes more


apparent when rays enter the crystal at an oblique angle to the face of the crystal
(e.g. near the radial edge of a system comprising a ring of detectors). In that case,
the rays can completely pass through the entrance crystal before interacting in
a neighbouring crystal. The ray is then mis-positioned as though it had entered
the neighbouring crystal or in some intermediate location, depending on the
relative amounts of energy deposited by the two interactions.
Spatial resolution is also affected by the energy of the photon and, for
scintillation detectors, the efficiency of collection of the scintillation light by the
PMTs. The energy of the ray that is deposited in the crystal determines the
amplitude of the measured signal, which in turn defines how accurately it can be
localized in the detector. The spatial resolution measured in a given crystal with
Tc (140keV) is inferior compared to that which would be measured with a
511keV photon.
As will be discussed later, the spatial resolution can also depend on the
count rate or amount of activity in the scanner. As the count rate increases, there
is an increased chance that two events will be detected at the same time in nearby
locations in the detector. These events will pile up and appear as a single event
at an intermediate location with a summed energy. This can lead to a loss of
resolution with increasing activity.
8.3.2. General measures of spatial resolution
There are several ways to characterize the spatial resolution, whether of a
detector or of a complete system. The point spread function (PSF) and line spread
function (LSF) are the profiles of measured counts as a function of position
across the point/line source. Rather than showing the complete profiles, however,
it is more convenient to characterize them by simple measures. The full width at
half maximum (FWHM) and full width at tenth maximum (FWTM) are useful
to describe the widths of the profile although they do not give information about
any asymmetry in the response. The equivalent width was defined as a way to
combine the FWHM and FWTM into a single parameter and describe the shape
of the profile in a simple way; it is defined as the width of a box function with
a height equal to the maximum amplitude of the profile and an area equal to
the total number of counts in the profile above 1/20 of its maximum amplitude.
Reducing the PSF or LSF to a few parameters carries with it a loss of information
about the spatial response of the imager; for example, LSFs or PSFs can have
very different shapes and still have the same FWHM.
The modulation transfer function (MTF) is one way to more completely
characterize the ability of a system to reproduce spatial frequencies. The MTF
is calculated as the Fourier transform of the PSF and is a plot of the response of
a system to different spatial frequencies. High spatial frequencies correspond to


fine detail and sharp edges, while low spatial frequencies correspond to coarse
detail. The better the response at high frequencies, the smaller the structures
that can be resolved. A flat response across all spatial frequencies means that
the system most accurately reproduces the object. As it is difficult to compare
imaging performance based on the MTF, however, the FWHM and FWTM are
used to characterize spatial resolution.
8.3.3. Intrinsic measurement spatial resolution
The intrinsic spatial resolution is a measure of the resolution at the detector
level (or detector pair level for PET) without any collimation. It defines the
best possible resolution of the system, since later steps in the imaging hardware
degrade the resolution from the detector resolution. On gamma cameras, the
intrinsic resolution is determined using a bar phantom with narrow slits of
activity across the detector. On PET systems, the intrinsic resolution is measured
as a source is moved between a pair of detectors operating in coincidence. The
FWHM and FWTM of profiles of detected counts as a function of position are
taken as measures of the intrinsic spatial resolution. In both cases, the intrinsic
spatial resolution sets a limit on the resolution but does not translate easily into
a clinically useful value because other components of the imager impact the
resolution in the image.
8.3.4. Extrinsic measurement spatial resolution
The spatial resolution of a nuclear medicine imager depends on many
factors other than just the detectors. The linear and angular sampling play a
significant role: to preserve the intrinsic resolution, the imager should be sampled
every 0.1FWHM. Under-sampling leads to small structures being missed in
the image. For single-photon imagers, a collimator is used to limit the direction
of rays incident on the detector. Collimators are designed for specific purposes
(e.g. sensitivity or resolution) and/or specific radionuclides. As the hole size and
spacing of a collimator will affect the spatial sampling, each collimator will lead
to different system spatial resolution.
The reconstruction processing performed to create tomographic images
in SPECT or PET also affects the image resolution. Reconstruction algorithms
are generally chosen to preserve as much fine detail and edge information as
possible, while keeping image noise sufficiently low so that it is not confused
with actual structure. The parameters of reconstruction can, therefore, change
with the imaging study and with the number of events measured.
The spatial resolution is not constant throughout the imaging field of view
(FOV). For PET systems, the resolution does not vary significantly with location


of the source between two detectors in a detector pair, but the systems radial
resolution often degrades as the source is moved radially outwards from the
centre of the scanner. For gamma cameras, the resolution degrades as the source
is moved away from the detector face. For this reason, system spatial resolution
measurements are performed with the source at different locations in the imaging
Extrinsic measures of spatial resolution are made under more clinically
realistic conditions and include the effects of the collimator (for single photon
imaging) and reconstruction processing. The extrinsic spatial resolution is
typically measured with a small point or line source of activity of a sufficiently low
amount such that effects seen at high count rates (i.e. mis-positioning of events)
are negligible. Measurements of system spatial resolution can be performed
in air or with scattering material added. A stationary source is positioned at
specified locations throughout the nuclear medicine imagers FOV. The spatial
resolution is determined from the images, including any reconstruction or
processing steps, by drawing profiles through the source. No spatial smoothing
or other post-processing is performed. In addition, any resolution modelling or
resolution recovery techniques applied during clinical reconstruction are not
used in the measurement of extrinsic resolution. The extrinsic spatial resolution
is distinguished from the intrinsic resolution because it includes many effects
not seen with the intrinsic resolution: collimator blurring, linear and angular
sampling, reconstruction algorithm, spatial smoothing, and impact of electronics.
While the extrinsic resolution measurement reflects the resolution of the
complete imaging system, the spatial resolution achieved in patient images
is typically somewhat worse than the extrinsic spatial resolution. The spatial
sampling is finer than occurs clinically because the pixel size is typically smaller
than that used for patient studies in order to sample the PSF or LSF sufficiently.
For imagers that reconstruct the data, the reconstruction algorithm in the
performance measurement is often not the technique applied to clinical data;
an analytical algorithm such as filtered back projection is generally specified
for tomographic systems to standardize results between systems. Another key
determinant of the clinical resolution is noise in the data that necessitates noise
reduction through spatial averaging (smoothing), which blurs the image. For
data with high statistics, a sharp reconstruction algorithm can be applied, and the
resulting image has good spatial resolution. For more typical nuclear medicine
studies, where the number of detected events is limited, some form of spatial
smoothing is applied, with the resulting blurring of fine structures.




8.4.1. Intrinsic measurement temporal resolution
As the activity in the FOV increases, events arrive closer to each other in
time until the imager cannot distinguish individual events. The timing resolution,
or resolving time, is the time needed between successive interactions in the
detector for the two events to be recorded separately. The timing resolution is
largely limited by the decay time of the crystal. For scintillators, the decay time
can be as high as 250300 ns or as low as 2040 ns, depending on the detector
material. Typically, the scintillation light does not decay with a single time
constant but with a combination of fast (nanosecond) and slow (microsecond)
components. For semiconductor detectors, the decay time is much smaller. In
addition to the detector decay time, the various components of the electronics can
contribute to the loss of temporal resolution. The timing resolution is generally of
less interest than its impact on the count rate performance of the system.
The timing of events is critical for PET, where two annihilation photons
must be detected within a timing window to be recorded as a valid event. The
timing window must be set wide enough to measure valid coincidence events but
not so wide that many coincidences between uncorrelated annihilation photons
(random coincidences) are accepted. Coincidence timing electronics are
carefully designed so that a detectors signal is processed as quickly as possible,
rather than waiting for the entire scintillation light to be measured. This allows
the coincidence timing window to be set to <10 ns, limited by the time of flight
of the two annihilation photons across the imagers diameter. For a ring diameter
of 90cm, theminimum coincidence time window would be 6 ns.
Recent developments in PET technology allow for the difference in times
of arrival (time of flight) of the two annihilation photons to be measured. For
time of flight systems, the coincidence time window is still 46 ns but the time
of flight difference can be measured with a resolution of 300600 ps. This time
of flight information is used in reconstruction to localize the events. The timing
resolution is measured with a low-activity source of activity by recording a
histogram of the number of events as a function of time difference.
8.4.2. Dead time
The consequence of a finite timing resolution is a loss of counts measured
at higher activities. When two photons arrive within the resolving time of the
detector, the two photons are seen by the electronics as a single event. One or both
of the events may be lost, and the events are also mis-positioned in space. The
random nature of radioactive decay means that there is always a possibility that


two events will arrive within the resolving time of the detector; this possibility
increases as the activity in the imager increases.
There are two kinds of dead time: non-paralysable and paralysable (seealso
Chapter6). Non-paralysable dead time arises when an event causes the system
to be unresponsive for a period of time, so that any later events that arrive
during that time are not recorded. For paralysable dead time, the second event
is not only not recorded but also extends the period for which the electronics
are unresponsive. At moderate count rates, paralysable and non-paralysable dead
times are the same; it is only at high count rates that the two types of dead time
differ (seeFig.8.2). It can be seen that systems with non-paralysable dead time
saturate at high count rates, while those with paralysable dead time peak and then
record fewer events as the activity increases. This leads to an ambiguity in the
measured count rate: the same observed count rate corresponds to two different
activity levels. The system dead time performance of nuclear medicine scanners
is typically intermediate between paralysable and non-paralysable dead time
because some components have paralysable dead time while other components
have non-paralysable dead time.

Measured count rate

No dead time

dead time

dead time

True count rate

FIG.8.2. System dead time as a function of count rate.



With increased dead time, additional activity injected in the patient does not
lead to a comparable improvement in image quality or reduction in image noise.
Dead time losses depend on the single event rate, coincidence count rate (for
PET), and the analogue and digital design characteristics of the nuclear medicine
imager. Dead time losses can depend on the activity distribution, especially for
PET because of the different single photon and coincidence rate relationship
with source distribution. They also depend on the radioisotope because dead
time results from all rays that interact in the detector, not just the photons that
fall within the energy window. For imaging studies with a large dynamic range
(e.g. cardiac scans), count rate performance is critical.
To correct for event losses due to dead time, a correction based on a
decaying source study is often applied to clinical data. The dead time correction
will generally correct for the loss of counts, so that the number of counts in the
image is independent of the count rate; it does not, however, compensate for the
higher image noise that arises because fewer events are actually measured.
8.4.3. Count rate performance measures
The generic measurement of count rate performance involves determining
the response of the nuclear medicine imager as a function of activity presented
to the system. Typically, this requires starting with a high amount of activity and
acquiring multiple images over time as the activity decays. The energy window is
set at low activity levels and is not changed at higher activities to accommodate
a shift in the photopeak due to pile-up effects. By comparing the observed events
with the counts that would be expected after decay correction of events detected
at low activities, the system dead time can be determined as a function of activity
level. It is especially important to determine the maximum measurable count
rate, since higher activities would result in no increase and perhaps a decrease
in detected counts. While most count rate performance measures call for starting
with a high activity and imaging as the activity decays, if too high an activity
is used at the beginning of the measurement, the detector may show effects of
saturation during later measurements at lower activities. Therefore, the amount
of activity at the beginning of the study must be sufficient to measure the peak
count rate but not be so high as to saturate the system for a significant period.
Intrinsic count rate performance measurements are performed with a source
in air and without any detector collimation. This is typically performed only on
gamma cameras. The system, or extrinsic, count rate performance is measured
with the complete system, including any collimation or detector motion, and a
distributed source with scattering material (e.g. a cylindrical phantom of specified
dimensions or a source placed within scattering material). The scatter adds low



energy photons that contribute to pile-up and dead time that are not present in the
intrinsic measurement.
For PET, random coincidences also increase as the activity increases;
whereas the true coincidence rate would increase linearly with activity in the
absence of dead time losses, the random coincidence rate increases quadratically
with activity, so that their impact becomes greater at higher count rates. The
activity where the random rate equals the true event rate is of importance, in
addition to the activity and count rate at which the true count rate saturates or
peaks. A global measure of the impact of random coincidences and scatter on
image quality is given in the noise equivalent count rate (NECR) defined as:

T + S + kR

where T, S and R are the true, scatter and random coincidence count rates,
respectively, and k is a factor that is equal to one if a smooth estimate of random
coincidences is used and two if a noisy estimate is used. This parameter does not
include reconstruction effects or local image noise differences but can be useful
in determining optimal activity ranges.
For systems that correct for dead time, it is important to apply dead time
correction and to reconstruct the data in addition to looking at the count rates.
The quantitative accuracy of the dead time correction is determined by looking at
a large region of interest in decay-corrected, reconstructed images; the counts in
the region of interest should be independent of activity level. It is also important
to examine the images at high activities for artefacts that may arise due to
spatially-varying mis-positioning effects or inaccuracies in various corrections
with increased activity.
8.5.1. Image noise and sensitivity
Images from nuclear medicine devices are typically noisy because the
amount of activity that can be safely injected and/or the scan duration without
patient discomfort or physiological changes in activity distribution is limited. The
number of detected events for a given amount of activity in the imaging systems
FOV is an important performance characteristic because a more efficient imager
can achieve low image noise with lower injected activity than a less efficient
system. Noise in the image can affect both visual (qualitative) image quality


and quantitative accuracy, especially in areas of low uptake or low contrast.

The relative response of a system to a given amount of activity is reflected in its
The sensitivity of a system is determined by many factors. The geometry of
the imager, especially the solid angle of the detectors, as well as any collimation
will determine how many photons reach the detectors. The stopping power and
depth of the detectors will impact how many of these photons are detected. In
addition, the radionuclides energy, coupled with the imagers energy resolution
and energy window, affect the number of accepted events. Finally, the number
of counts measured in a given time for a fixed amount of activity depends on the
source distribution and its position in the imager.
8.5.2. Extrinsic measure sensitivity
All performance measurements of sensitivity are extrinsic; for single
photon imaging, in particular, the collimator is a major source of loss of events,
so it is more clinically interesting to know the sensitivity of the system with a
particular collimator.
As noted above, the number of observed counts depends greatly on
the activity distribution. For this reason, any measurement of sensitivity is
performed under prescribed conditions that do not attempt to replicate patient
activity distributions. The source configurations and definitions of sensitivity
vary widely, however. For planar imaging, a shallow dish source without
intervening scatter material is used, and the sensitivity is reported as a count rate
per activity. For SPECT, a cylindrical phantom is filled uniformly with a known
activity concentration, and the sensitivity is reported as a count rate per activity
concentration. For whole body PET scanners, a line source that extends through
the axial FOV is imaged with sequentially thicker sleeves of absorbing material,
and the data are extrapolated to the count rate one would measure without any
absorber; the sensitivity is then reported as a count rate per unit activity. Small
animal PET systems use a point source in air centred in the scanner, and the count
rate per activity, as well as the absolute sensitivity (in per cent) are reported.
None of these sensitivity measurements can be used to predict the number of
events that will be observed for patient studies; however, systems with higher
sensitivity will generally record more events from a patient activity distribution
than those with lower sensitivity.




8.6.1. Image uniformity
The uniformity of response of a nuclear medicine imager across the FOV is
important for both qualitative and quantitative image quality. All PMTs of a given
type do not respond exactly the same way, and a correction for this difference in
gain is applied before the image is formed. Collimators can also have defects that
lead to non-uniformities in the image. For tomographic scanners, corrections for
attenuation and unwanted events such as scatter can also affect the uniformity of
the image.
Intrinsic uniformity is measured without a collimator by exposing the
detector to a uniform activity distribution (e.g. from a distant, uncollimated
point source). Intrinsic uniformity is measured at both low and high count rates,
where mis-positioning effects become more pronounced. The extrinsic system
uniformity is determined with a collimator in place (for single photon imaging),
and images are processed or reconstructed as for clinical studies. In both cases,
sufficient counts must be detected, so that image noise is low. Quantitative
assessment of image uniformity includes variation of pixel counts in small
regions across the FOV. However, because simple metrics of non-uniformity
such as this do not provide a complete assessment of what the eye perceives in
the image, a visual analysis is also important.
8.6.2. Resolution/noise trade-off
Most performance measurements are carried out under non-clinical
conditions to isolate an aspect of the imagers performance. To include more
of the effects seen in clinical data, some performance standards call for a
measurement of image quality. The activity distribution is a series of small
structures (e.g. spheres of varying diameters) in a background activity typical of
the activity levels seen in patient studies. The activity is imaged for a clinically
relevant time, so that the noise level in the data is comparable to that in typical
patient studies. The data are processed in the same manner as clinical data. The
resulting image, then, is a better representation of the resolution and noise seen
clinically. Data analysis consists of such measures as sphere to background
contrast recovery, noise in background areas and/or signal to noise ratio in the
spheres. While still a simplistic and non-clinical distribution, the measurement
gives a more relevant indication of clinical resolution/noise performance.




There are many other performance measures that reflect a given aspect of
a nuclear medicine imager. For planar systems, the spatial linearity, or spatial
distortion of the measured position of photons compared to the actual position, is
important for good image quality. A number of nuclear medicine imaging systems
incorporate anatomical (e.g. computed tomography or magnetic resonance
imaging) imagers into the scanner, and the images from the different modalities
must be registered spatially. Another area where spatial registration is necessary
is in single photon systems where multiple energy windows are used, and the
images acquired in the different windows must be overlaid to form the image.
Quantitative linearity and calibration is an important measurement for systems
such as PET scanners that aim to relate pixel values to activity concentrations.


Department of Nuclear Medicine,
St. George Hospital,
Sydney, Australia
9.1.1. Construction of dose calibrators
Throughout the world, the instrument that is used in nuclear medicine to
measure radioactivity is the calibrated re-entrant ionization chamber, commonly
known as a radionuclide calibrator or dose calibrator. Commercial systems
comprise a cylindrical well ionization chamber connected to a microprocessorcontrolled electrometer providing calibrated measurements for a range of common
radionuclides (Fig. 9.1). The chamber is usually constructed of aluminium filled
with argon under pressure (typically 12 MPa or 1020 atm). Dose calibrators
with reduced gas pressure are available for positron emission tomography (PET)
production facilities where very large activities may be measured.

FIG.9.1. A typical dose calibrator (e.g. CRC 25R).



A well liner, made of low atomic number material (e.g. lucite (Perspex))
which can be removed for cleaning, prevents the ionization chamber from
becoming accidentally contaminated. A sample holder is provided into which
a vial or syringe can be placed to ensure that it is positioned optimally within
the chamber. The dose calibrator may include a printer to document the activity
measurements or an RS-232 serial communications port or USB port to interface
the calibrator to radiopharmacy computerized management systems.
The chamber is typically shielded by the manufacturer with 6 mm of
lead to ensure low background readings. Depending on the location of the dose
calibrator, the user may require additional shielding, either to reduce background
in the chamber or to protect the operator when measuring radionuclides of
high energy and activity. However, this will alter the calibration factors due to
backscattering of photons together with the emission of Pb K shell X rays arising
from interactions within the lead shielding. If additional shielding is used, the
dose calibrator should be recalibrated or correction factors determined to ensure
that the activity readings remain correct.
As examples of commercial systems, the specifications of two widely used
dose calibrators are given in Table 9.1.




Capintec CRC-25R


Atomlab 200

Ionization chamber

26 cm deep 6 cm diameter

26.7 cm deep 7 cm diameter

Measurement range

Autoranging from 0.001 MBq

to 250 GBq

Autoranging from 0.001 MBq

to 399.9 GBq

Nuclide selection

8 pre-set, 5 user-defined
(80 radionuclide calibrations in

10 pre-set, 3 user-defined
(94 radionuclide calibrations in

Display units

Bq or Ci

Bq or Ci

Electrometer accuracy



Response time

Within 2 s

1 s for activities >75 MBq






9.1.2. Calibration of dose calibrators

A dose calibrator can be calibrated in terms of activity by comparison with
an appropriate activity standard that is directly traceable to a national primary
standard. National primary standards are maintained by the relevant national
metrology institute, such as the National Physical Laboratory (NPL) in the United
Kingdom, the National Institute of Standards and Technology in the United States
of America and the Australian Nuclear Science and Technology Organisation
(ANSTO). Using the activity standard, a calibration factor for the ionization
chamber can be determined for the specific radionuclide. The reciprocal of the
calibration factor represents the efficiency N of the ionization chamber for the
radionuclide N.
The nuclide efficiency N can be expressed as the sum of two components:
N =

p i (E i ) i (E i )


pi(Ei) is the emission probability per decay of photons of energy Ei;
and i(Ei) is the energy dependent photon efficiency of the ionization chamber.
Figure 9.2 illustrates a typical efficiency curve as a function of photon
energy. Thin-walled aluminium chambers show a strong peak in efficiency
at photon energies around 50 keV. This results from the rapid increase of the
probability of photoelectric interactions in the filling gas with decreasing energy
and the low energy cut-off with aluminium walls at about 20 keV.
Knowing the energy dependent photon efficiency curve for a specific
ionization chamber will enable the nuclide efficiency for any radionuclide to be
determined from the photon emission probability for each photon in its decay.
The 511 keV annihilation radiation will be measured when the activity of
positron emitting radionuclides is to be assayed. A single calibration factor for all
positron emitters cannot be used as the emission probability of the positrons must
be taken into account. The probability (branching ratio) of positron emission for
C is 100% and for 18F is 96.7%.



FIG.9.2. Efficiency curve as a function of photon energy.

9.1.3. Uncertainty of activity measurements

The following sections describe the major sources of uncertainty in dose
calibrator measurements. Calibration factor
For medical radionuclides, such as 99mTc and 131I, the uncertainty of
national standards is typically in the range of 13%. However, when the standard
is used to calibrate a medical dose calibrator, the uncertainty will be larger due
to the inherent limit on instrument repeatability. Furthermore, the calibration
factor will be for the particular vial size and thickness, and volume of solution,
used for the national standard. The calibration factor for a different container
(a syringe) and/or a different volume may vary from the established calibration
by a significant amount (see Section Electronics
Electrometers measure the current output from the ionization chamber
ranging from tens of femtoamperes up to microamperes a dynamic range
of 108, corresponding to activity levels from kilobecquerels to hundreds of
gigabecquerels. Modern dose calibrators automatically adjust the range while
older units required the operator to select the appropriate range. The potential
for different linearity characteristics for each range may result in discontinuities
when the range is changed. The effects of inherent inaccuracy, linearity and range
changing are illustrated in Fig. 9.3. The linearity of the dose calibrator must be


established over the full range of intended use when the unit is commissioned
and verified as part of the quality control programme (see Section

FIG.9.3. Electrometer inaccuracies (courtesy of the National Physical Laboratory). Statistical considerations

Repeated measurements on a single sample will not be identical because
of the random nature of radioactive decay (see Chapter 5). If the measurement
period remains constant, the precision of the measured activity will increase as
the activity increases. Conversely, the precision will deteriorate for low activity
sources. To compensate for this, many calibrators automatically adjust the
measurement period depending on the activity level. This may vary from less
than one second to tens of seconds for low activities (<1 MBq). Ion recombination
As the activity of the source increases, the probability of recombination
of the positive ions with electrons increases. At high source activities, this can
become significant and lead to a reduction in the measured current. The effect of
recombination is illustrated in Fig. 9.4. For most modern calibrators, the effects
of recombination should be less than 1% when measuring 100 GBq of 99mTc. Background radiation
When the source holder is empty, the dose calibrator will still record
a non-zero reading due to background radiation. This will comprise natural
background and background from sources within the radiopharmacy. It could


also be due to contamination on either the source holder itself or the well liner.
Most dose calibrators provide a background subtraction feature. An accurate
measurement of the existing radiation level is made by the calibrator (usually
integrating over several minutes to improve precision) which is then automatically
subtracted from each subsequent reading. This may lead to erroneous results if
the background radiation has changed since it was measured due to the presence
of additional nearby sources or contamination. It is, therefore, essential to make
regular checks of the background radiation level.

FIG.9.4. Effects of recombination (courtesy of the National Physical Laboratory). Source container and volume effects

Variations in the composition and thickness of the source container will
give rise to corresponding variations in the measured activity. These effects
will be most noticeable for low energy photon emitters and pure emitters.
Measurements made at NPL, United Kingdom (Table 9.2) have shown that
variations in glass wall thicknesses, which were within the range of the vial
manufacturing tolerances, could lead to errors of up to 7% for 125I.
When the activity is drawn into a syringe, the source geometry will be
different from that in a vial. Not only will the composition and thickness of the
syringe wall be different from that of the vial, but the distribution of the source
will also be different depending on the size of syringe used. This is clearly evident
in Fig. 9.5, showing measurements at NPL for 111In in three sizes of syringe (1, 2
and 5 mL) from two different manufacturers in comparison to those measured in
a laboratory standard P6 vial. Also illustrated in Fig. 9.5 is the effect of changing
the source volume without changing the activity. Self-absorption of the emitted




Reduction in response with increase in vial wall thickness of

0.08 mm

0.2 mm














mL Sabre
mL Sabre
mL Sabre
mL Gillette
mL Gillette
mL Sabre

FIG.9.5. The effects of geometry and sample size on dose calibrator readings, demonstrated
for 111In measured in varying syringes (reproduced from Ref. [9.1]).

radiation will change as the source volume changes. This will be particularly
important for radionuclides with low energy components such as 123I. For 99mTc,
the correction will usually be less than 1% but should be confirmed for a new
dose calibrator or when the supplier of the syringes changes. Source position
The manufacturers source holder is designed to keep the source at the area
of maximum response on the vertical axis of the well. Variations in response
due to changes in vertical height or horizontal position of a few millimetres are
usually insignificant.


CHAPTER 9 Source adsorption

Certain radiopharmaceuticals have been observed to adsorb to the surface
of the container. For example, up to 30% of the activity of 201Tl has been
found to be adsorbed onto the glass of P6 vials. 99mTc-tetrofosmin has been
shown to adsorb onto the surface of syringes, such that some types of syringe
may retain as much as 19% of the activity. Of this, 6% adhered to the rubber
plunger with the remainder attached to the plastic syringe barrel. Up to 15% of
Tc-macroaggregate of albumin (MAA) may adhere to the syringe, although
the amount on the rubber plunger is usually no more than 1%. The possibility of
activity adsorption should be considered whenever the facility uses syringes from
a different manufacturer.
9.1.4. Measuring pure emitters
The detection efficiency of ionization chambers for radiation is
low as most, if not all, of the particles are absorbed in the source solution
(self-absorption), in the walls of the container or in the walls of the ionization
chamber. The dose calibrator response from particles will be almost entirely
from bremsstrahlung radiation (see Section 1.1.7). In the energy region of interest
for ionization chamber measurements, the bremsstrahlung photon spectrum
is roughly the same shape as the particle energy distribution. The average
particle energy is, therefore, a good parameter with which to characterize the
ionization chamber response to the bremsstrahlung radiation.
Bremsstrahlung radiation flux is proportional to the square of the atomic
number of the absorbing material. Thus, in argon-filled ionization chambers,
significant activities are required in order to obtain a precise estimate of the
activity. However, as substantial activities of radionuclides are required to be used
therapeutically, reliable measurements are possible using pure emitters used
clinically such as 90Y, 89Sr and 32P. However, geometry factors (see Section
will be even more important and the system must be calibrated for the specific
containers and volumes to be used clinically. Manufacturers are now producing
dose calibrators specifically for the measurement of emitters. These use a sodium
iodide detector instead of an ionization chamber, resulting in a significantly
increased detection efficiency; however, as the manufacturers state in their product
literature, measurements still require exacting attention to the sample container,
the sample volume and activity concentration to achieve accurate results.
Most commercially available ionization chambers are provided with
calibration factors for commonly used emitters, although these will usually
correspond to the activity within a vial rather than a syringe. The type of vial used
in the calibration is often unspecified, so the user should verify the calibration in


the vials normally used in the practice. Similarly, the calibration of the activity
within the size of syringe to be used clinically should be established. Published
results comparing the intrinsic efficiencies of dose calibrators from five different
manufacturers found that all systems had a good calibration for 32P, a reduction
in efficiency of approximately 1020% for 89Sr, and a wide divergence in
efficiency for 90Y. For this radionuclide, the results obtained using the calibration
factors supplied by the manufacturers ranged from 64 to 144% of the true value,
re-emphasizing the need for the calibration to be confirmed within the nuclear
medicine department.
Several emitters used for radionuclide therapy include a ray component.
These radionuclides include 131I (364 keV, 81.5% abundance) and 186Re (137 keV,
9.5% abundance). For these radionuclides, the ionization chamber efficiency
is primarily determined by the contribution and the manufacturers supplied
calibrations will usually be accurate to within 10%.
9.1.5. Problems arising from radionuclide contaminants
Unfortunately, it is often not possible for a solution of a radionuclide to
be totally free of other radionuclides. The proportion of the total radioactivity
that is present as a specific radionuclide is defined as the radionuclide purity.
National and international pharmacopoeia specify the radionuclidic purity of
a radiopharmaceutical. For example, the European Pharmacopoeia entry for
Ga-citrate injection requires that no more than 0.2% of the total radioactivity
be due to 66Ga. This requirement must be met at all times up to the expiry time of
the product. The US Pharmacopoeia is less stringent, specifying that not less than
99% of the total radioactivity be present as 67Ga at the time of calibration.
The presence of contaminants, even when less than 1% of the total activity,
can have a marked effect on the ionization chamber current and, thus, on the
measured activity. The British Pharmacopoeia specification for 201Tl-thallous
chloride requires that Not more than 2.0 percent of the total radioactivity
is due to thallium-202 and not less than 97.0 percent is due to thallium-201.
Thallium-202 has a half-life of 12.2 d and the predominant photon energy is
440 keV. Another possible contaminant is 200Tl which has a half-life of 1.09 d
and prominent energies at 368 keV and 1.2 MeV. Both of these radionuclide
contaminants will have a high efficiency in a dose calibrator. As the half-life of
Tl is significantly longer than that of 201Tl, the relative proportion of 202Tl to
Tl will increase over time. If the accuracy of a dose calibrator is to be checked
with a 201Tl source, the apparent accuracy could change depending on when the
measurements are taken relative to the stated calibration date. The presence of
these high energy contaminants will have an adverse effect on image quality due
to increased septal penetration and will also lead to an increased radiation dose to


the patient. The effective dose, in millisieverts per megabecquerel, for 200Tl, 201Tl
and 202Tl is 0.238, 0.149 and 0.608, respectively. It should be noted that these
problems will be increased if the radiopharmaceutical is administered prior to the
nominal calibration date, as the proportion of 200Tl will be higher.
9.2.1. Acceptance tests
Acceptance tests for dose calibrators should include measurements of the
accuracy, reproducibility, linearity and geometry response. These are required to
ensure that the unit meets the manufacturers specifications and to give baseline
figures for subsequent quality control. Accuracy and reproducibility
The accuracy is determined by comparing activity measurements using
a traceable calibrated standard with the suppliers stated activity, corrected for
radioactive decay. The accuracy is expressed in per cent deviation from the actual
activity and should be measured for all radionuclides to be used routinely. It is
recommended that measurements of a long lived source, for example 137Cs, be
recorded at the time of initial testing for each radionuclide setting to be used
clinically for later quality control.
The reproducibility, or constancy, can be assessed by taking repeated
measurements of the same source. If the sample holder is removed from the
chamber between each measurement, the measured reproducibility will include
any errors associated with possible variations in source position. Linearity
There are several approaches to the measurement of the linearity response of
a dose calibrator. Typically, a vial containing a high activity of 99mTc is measured
repeatedly over a period of at least 5 d. During this time, a 100 GBq source will
decay to 0.1 MBq. It is essential that the initial activity represents the highest
activity that is likely to be used in clinical practice, which will usually be the
first elution from a new Mo/Tc generator. A semi-log plot of the measurements,
corrected for background, should follow the expected decay of the radionuclide.
Any deviation from the expected line at high activities indicates saturation of
response of the ionization chamber. Accurate background measurements, at the


time of each assay, are essential as the background will become an increasing
component of the reading as the source decays. Deviations from linearity at low
activities are likely to be due to radionuclide impurities, such as 99Mo in vials
containing 99mTc.
Another approach that can be used to check the linearity requires a series of
radioactive sources that cover the range of activities to be measured. The sources
should all be prepared from the same stock solution and the dispensed volumes
measured accurately by weighing the vials pre- and post-dispensing. The volume
of liquid in each vial should be adjusted with a non-radioactive solution, so that
the volume is identical in each vial, to eliminate any geometry dependency in
the measurement. The measured activities are corrected for decay to the time of
measurement of the first vial and plotted against the dispensed volumes to assess
the calibrator linearity. The error in this method will be increased if there are
any small variations in the vial wall thickness as the same vial is not used for all
Finally, linearity can be assessed by repeated measurements on a single
vial using a series of graded attenuators appropriate for a specified test source to
reduce the measured ionization current. These are typically a series of concentric
cylinders that fit over the vial. The attenuation through each cylinder must be
accurately known to use this method. Geometry
The measured activity may vary with the position of the source within
the ionization chamber, with the composition of the vial or syringe, or with the
volume of liquid within the vial or syringe. Appropriate correction factors must be
established for the containers and radionuclides to be used clinically, especially
if radionuclides that have a substantial component of low energy photons, such
as 123I, are to be used. For each vial or syringe to be used clinically, a series of
measurements should be undertaken in which the activity remains constant, but
the volume is increased from 10 to 90% of the maximum volume by the addition
of water or saline. Corrected for decay, a plot of activity against volume should
be a straight horizontal line. Any deviations from this can be used to calculate the
appropriate correction factor.
Similarly, vial to syringe correction factors can be determined by measuring
the activity transferred from the vial to the syringe (original vial activity minus
residual activity) and comparing this to the activity measured in the syringe itself.
Geometry dependencies should not change over time; however, if
the practitioner changes the manufacturer of the syringes or obtains the
radiopharmaceuticals in a different vial size, a new set of calibration factors
should be determined.


9.2.2. Quality control Background check
As noted in Section, when the source holder is empty, the dose
calibrator will still record an activity due to background radiation. This
will come from natural background, from sources within the radiopharmacy
and/or from contamination present on the source holder or well liner. It is a
useful practice to keep a spare source holder and a spare well liner, so that if
contamination is detected the contaminated item can be removed from service to
be decontaminated, or left until the radioactivity has decayed.
At a minimum, the background should be determined each morning before
the dose calibrator is used and recorded. The background subtraction feature,
if available, can be used at that time to remove the measured background from
subsequent measurements. The technologist should also confirm the absence of
any additional background before all activity measurements during the day. Check source reproducibility
A long lived check source should be used on a daily basis to confirm the
constancy of the response of the dose calibrator. Sealed radioactive sources
of 57Co and 137Cs, shaped to mimic a vial, are available commercially for this
purpose. The check source should be measured on all radionuclide settings that
are used clinically. Although the recorded activity of a 137Cs source on the 99mTc
setting will not be a correct measurement of its activity, a reading outside of that
expected from previous results may indicate a faulty dose calibrator or a change
in calibration factor, in this case of 99mTc.
The International Electrotechnical Commission (IEC) has published two
standards [9.2, 9.3] and a technical report [9.4] relating to dose calibrators. IEC
standards are often adopted by national standards organizations. Reference [9.3]
is for manufacturers to use to ensure that the equipment performance is specified
in a standardized way, while Ref. [9.4] is aimed at the users of dose calibrators.
There should also be national standards covering dose calibrators. The
American National Standards Institute publication ANSI N42.13-2004 [9.5] is
often referenced by US manufacturers. This specifies the minimum requirements
in terms of accuracy and reproducibility for dose calibrators:



The accuracy of the instruments, at activity levels above 3.7 MBq shall be
such that the measured activity of a standard source shall be within 10%
of the stated activity of that source [9.5];
The reproducibilityshall be such that all of the results in a series
of ten consecutive measurements on a source of greater than 100 Ci
(3.7 106 Bq) in the same geometry shall be within 5% of the average
measured activity for that source [9.5].
National metrology institutes are responsible for the development and
maintenance of standards, including activity standards. These institutes, often
in collaboration with the relevant national professional body, have undertaken
national comparisons of the accuracy of the dose calibrators used in clinical
practice. Such comparisons have used, where possible, the clinical radionuclides
Ga, 123I, 131I, 99mTc and 201Tl, and have been carried out in Argentina, Australia,
Brazil, Cuba, the Czech Republic, Germany, India and the United Kingdom.
In some countries, such as Cuba and the Czech Republic, participation in the
comparison is mandatory, while in many other countries it is voluntary. The
surveys can also be used to measure the reproducibility of the calibrators.
As an example, Table 9.3 shows the results from a survey undertaken in
Australia in 2007.








No. of calibrators





Within 5% error





Within 10% error





Within 10%





These surveys also offer the opportunity for the calibration factor to be
adjusted if a dose calibrator is found to be operating with an error of >10%.




9.5.1. Adjusting the activity for differences in patient size and weight
Protocols used in nuclear medicine practices should specify the usual
activity of the radiopharmaceutical to be administered to a standard patient. In
most western countries, the standard patient is taken to be one whose weight is in
the range 7080 kg. However, many patients fall outside of this range. If a fixed
activity is used for all patients, this will lead to an unnecessarily high radiation
exposure to an underweight patient and may lead to images of unacceptable
quality or very long imaging times in obese patients.
There have been various approaches to determining the activity to be
administered. These are usually designed to provide a constant count density in
the image to maintain image quality or to provide a constant effective dose to the
patient. For example, it has been shown that for myocardial perfusion scans using
Tc-tetrofosmin, the activity should be increased by 150% for a 110 kg patient
and by 200% for a 140 kg patient in order to maintain image quality without
increasing imaging time.
It has been shown, using the radiation dose tables provided in International
Commission on Radiological Protection (ICRP) publications 53, 80 and 106
[9.69.8], that the effective dose (mSv/MBq) can be expressed as a simple
power function of body weight. Scaling factors for the activity, to give a constant
effective dose can, therefore, be derived from the expression (W/70)a, where W
represents the weight of the person and the power factor a is specific for the
radiopharmaceutical. Again, using 99mTc-tetrofosmin as an example, a is found
to be 0.834. Although the dosimetry models are only available up to 70 kg, this
power function can be extrapolated to derive scaling factors for patients whose
weight exceeds 70 kg. Using this approach, the activity should be increased by
146% for a 110 kg patient and by 178% for a 140 kg patient. This approach is
useful, but should be used with caution. The extrapolated activity would lead to
comparable organ and tissue doses for a patient of large body build but not for
a patient of similar weight due to large body fat deposits as the biodistribution
of the radiopharmaceutical would not be the same in these two cases. Table 9.4
presents the a value for common radiopharmaceuticals.
9.5.2. Paediatric dosage charts
Children are approximately three times more radiosensitive than adults, so
determining the appropriate activity to be administered for paediatric procedures
is essential. In addition to the scaling factor to be applied to the adult activity, a




a value


















a value







Tc-red cells


Tc-white cells












I or 131I iodide



minimum activity must be specified in order to ensure adequate image quality.

In the past, the scaling factors were assessed using weight alone or body surface
area obtained from both height and weight. These two methods can give rise to
quite different scaling factors. For example, the scaling factor for a 20 kg child
is 29% of the adult activity using weight alone, but 43% when based on body
surface area.
Recently, the European Association of Nuclear Medicine (EANM)
Dosimetry and Paediatric Committees have prepared a dosage card which
recognizes that a single scaling factor is not optimal for all radiopharmaceuticals.
They used the methodology presented in Section 9.5.1 and were able to establish
that radiopharmaceuticals could be grouped into three classes (renal, thyroid and
others), with different scaling factors for each class. A dosage card is available on
the EANM web site that gives the minimum recommended activity and a weight
dependent scaling factor for each radiopharmaceutical which was determined to
give weight independent effective doses. This dosage card is reproduced here as
Fig. 9.6. To assist in these calculations, an on-line dosage calculator is available
on the EANM web site1, in which the user specifies the childs weight and the
radiopharmaceutical, and the recommended activity is displayed.




9.5.3. Diagnostic reference levels in nuclear medicine

The recommendations of the ICRP specifically exclude medical exposures
from its system of dose limits, as the patient is directly benefiting from the
radiation exposure. However, in Publication 73 (1996) [9.9], the ICRP introduced
the term diagnostic reference level (DRL) for patients. DRLs are investigation
levels and are based on an easily measured quantity, usually the entrance surface
dose in the case of diagnostic radiology, or the administered activity in the case
of nuclear medicine. DRLs are referred to by the IAEA as guidance levels in
Safety Reports Series No. 40 [9.10], published in 2005. This publication contains
a table of guidance levels reflecting the values used in the early 1990s, when
single photon emission computed tomography procedures were far less common.
A survey of the use of DRLs in eight European countries, published in 2007
[9.11], showed that their introduction in nuclear medicine varied considerably.
For example, France had set DRLs for 10 nuclear medicine procedures, Germany
for 17 procedures, Italy for 48 procedures, while the United Kingdom listed 96
procedures. In some countries, the DRLs were set at the activities for which
marketing approval had been given, while in other countries the DRLs were
determined for each procedure by calculating the 75th percentile of the spread of
data values collected from a survey of participating practices. The latter approach
has been widely used to set DRLs in radiology and has been used in other parts
of the world, such as Australia and New Zealand, to establish DRLs in nuclear



Dosage Card

(Version 1.2.2014)
Multiple of Baseline Activity



























































































































A[MBq]Administered = BaselineActivity x Multiple

a) For a calculation of the administered activity, the baseline activity value has to be multiplied by the multiples given above for the recommended radiopharmaceutical class
(see reverse).
b) If the resulting activity is smaller than the minimum recommended activity, the minimum activity should be administered.
c) The national diagnostic reference levels should not be exceeded!
a) 18F FDP-PET Brain,
50 kg:

activity to be administered [MBq] = 14.0 x10.71 [MBq]

150 MBq

b) 123ImIBG,
3 kg:

activity to be administered [MBq] = 28.0 x1 [MBq] = 28 MBq

< 37 MBq (Minimum Recommended Activity)
activity to be administered: 37 MBq

This card is based upon the publication by Jacobs F, Thierens H, Piepsz A, Bacher K, Van de
Wiele C, Ham H, Dierckx RA. Optimized tracer-dependent dosage cards to obtain weightindependent effective doses. Eur J Nucl Med Mol Imaging. 2005 May; 32(5):581-8.
This card summarizes the views of the Paediatric and Dosimetry Committees of the EANM
and reflects recommendations for which the EANM cannot be held responsible.
The dosage recommendations should be taken in context of good practice of nuclear
medicine and do not substitute for national and international legal or regulatory provisions.

Android App

iPhone App

EANM Executive Secretariat

Hollandstrasse 14/Mezzanine 1020 Vienna, Austria
Phone: +43-1-2128030, fax: +43-1-21280309
office@eanm.org - www.eanm.org - fb/officialEANM

FIG.9.6. European Association of Nuclear Medicine (EANM) paediatric dosage card

(courtesy of EANM).



Recommended Amounts in MBq



Baseline Activity
(for calculation
purposes only)




I (Thyroid)


I Amphetamine (Brain)




I HIPPURAN (Abnormal renal function)



I HIPPURAN (Normal renal function)









F FDG-PET torso



F FDG-PET brain



F Sodium fluoride



Ga Citrate



Tc ALBUMIN (Cardiac)



Tc COLLOID (Gastric Reflux)



Tc COLLOID (Liver/Spleen)



Tc COLLOID (Marrow)






Tc DTPA (Abnormal renal function)



Tc DTPA (Normal renal function)



Tc ECD (Brain perfusion)



Tc HMPAO (Brain)






Tc IDA (Biliary)



Tc MAA / Microspheres









Tc Pertechnetate (Cystography)



Tc Pertechnetate (Ectopic Gastric Mucosa)



Tc Pertechnetate (Cardiac First Pass)



Tc Pertechnetate (Thyroid)




Tc RBC (Blood Pool)



Tc SestaMIBI/Tetrofosmin
(Cancer seeking agent)



Tc SestaMIBI/Tetrofosmin2
(Cardiac rest scan 2-day protocol min)



Tc SestaMIBI/Tetrofosmin2
(Cardiac rest scan 2-day protocol max)



Tc SestaMIBI/Tetrofosmin2
(Cardiac stress scan 2-day protocol min)



Tc SestaMIBI/Tetrofosmin2
(Cardiac stress scan 2-day protocol max)



Tc SestaMIBI/Tetrofosmin2
(Cardiac rest scan 1-day protocol)



Tc SestaMIBI/Tetrofosmin2
(Cardiac stress scan 1-day protocol)

















Tc Spleen (Denatured RBC)


Tc TECHNEGAS (Lung ventilation)3

The minimum recommended activities are calculated for commonly used gamma cameras or positron
emission tomographs. Lower activities could be administered when using systems with higher counting

The minimum and maximum values correspond to the recommended administered activities in the
EANM/ESC procedural guidelines (Hesse B, Tagil K, Cuocolo A, et al). EANM/ESC procedural guidelines for
myocardial perfusion imaging in nuclear Cardiology. Eur J Nucl Med Mol Imaging. 2005 Jul;32(7):855-97.

This is the activity load needed to prepare the Technegas device. The amount of inhaled activity will be lower.

FIG.9.6. European Association of Nuclear Medicine (EANM) paediatric dosage card

(courtesy of EANM) (cont.).




9.6.1. Surface contamination limits
Surface contamination with radioactivity could lead to contamination of a
radiation worker and/or external irradiation of the skin of the worker. Internal
contamination could arise from inhalation and/or ingestion of the radionuclide.
The surface contamination limits given in Table 9.5 were derived based on
a committed effective dose limit of 20 mSv/a and the models for inhalation
and ingestion given in ICRP publications 30, 60 and 61 [9.129.14]. For each
radionuclide, the most restrictive pathway (inhalation, ingestion or external
irradiation) was used.
Surfaces in designated
areas, including
protective clothing

Interiors of glove boxes

and fume cupboards

Non-designated areas
including personal
clothing (Bq/cm2)





















































9.6.2. Wipe tests and daily surveys

Surveys of the radiopharmacy must be undertaken to ensure that these
surface contamination limits are not exceeded and that the operator is not
unnecessarily exposed to external radiation. Exposure could result from sources
inadvertently left on a bench and from contamination on bench surfaces.
Aerosolized droplets from a syringe during dispensing may go unnoticed, so it is
essential that all staff are aware that the dispensing area may be contaminated and
always wear protective gloves when working in this area. All radiopharmaceutical
elution, preparation, assay and administration areas should be surveyed at the end
of each working day.
Surveys should initially be undertaken with a survey meter to ensure that
no unexpected exposed sources are present in the radiopharmacy. All surfaces
should then be checked for contamination using a contamination monitor with
a probe appropriate to the radionuclides used. The background radiation levels
in the radiopharmacy, particularly in the dispensing area, are often higher than
elsewhere in the nuclear medicine department, so quantifying any contamination
found using a probe is difficult. If a low energy emitter is being used, it will
prove difficult or impossible to detect with an external probe. In these situations,
a wipe test should be used. A minimum area of 100 cm2 should be wiped and
then the activity on the wipe can be assessed using a pancake probe, or more
accurately in a well counter. For low energy emitters, such as 3H or 14C, liquid
scintillation counting must be used. When quantifying the surface contamination,
it is generally assumed that a wipe test using a dry wipe will remove one tenth of
the contamination while a wet wipe will remove one fifth of the contamination.
9.6.3. Monitoring of staff finger doses during dispensing
Systematic studies of the dose to the hands of staff working in
radiopharmacies have shown that finger doses may approach or exceed the annual
dose limit of 500 mSv to the extremities. The most exposed parts of the hands
are likely to be the tips of the index and middle fingers, and the thumb of the
dominant hand, with exposure for the index finger being highest. The ICRP has
recommended that finger dose monitoring be undertaken for any person handling
more than 2 GBq/d and regular monitoring should be carried out if doses to the
most exposed part of the hand exceed 6 mSv/month.
Although the dose to the finger tip will be the highest, it is much more
practical to wear a ring monitor at the base of the finger. A thermoluminescent
dosimeter chip mounted in a plastic ring is usually the most convenient type of
monitor. Such monitors are often available in a variety of sizes. The ring should
fit tightly, so that it is not inadvertently removed when the gloves are taken off.


The ICRP recommends that the ring monitor be worn on the middle finger with
the element positioned on the palm side, and that a factor of three should be
applied to derive an estimate of the dose to the tip. If the dosimeter element is
worn facing towards the back of the hand, a factor of six should be applied.
The dose to the fingers is critically dependent on the dispensing technique
used and the skill of the operator. It is important that staff undertake extensive
training in the dispensing technique with non-radioactive solutions prior to
dispensing radiopharmaceuticals for the first time. This is particularly important
with PET radiopharmaceuticals as the specific dose rate constant is much higher
for positron emitters than for radionuclides used for single photon imaging.
9.7.1. Fume cupboards
A fume cupboard is an enclosed workplace designed to prevent the spread
of fumes to the operator and other persons. The fumes can be in the form of
gases, vapours, aerosols or particulate matter. The fume cupboard is designed
to provide operator protection rather than protection for the product within the
cabinet. A fume cupboard would, therefore, not be suitable as an area for cell
labelling procedures as this requires that the blood remain sterile at all times.
Fume cupboards usually include a transparent safety screen which can be
adjusted either vertically (more commonly) or horizontally to vary the size of the
working aperture into the cabinet. Some cupboards are available with a lead glass
safety screen to minimize the need for additional radiation shielding. The most
common type of fume cupboard is known as a variable exhaust air volume fume
cupboard which maintains a constant velocity of air into the cabinet (the face
velocity) irrespective of the sash position. Figure 9.7 shows a fume cupboard
suitable for use with radioactive materials.
Fume cupboards are available which discharge the exhaust air directly,
or after carbon filtration, to the atmosphere, usually above the building. Other
cabinets, known as recirculating fume cabinets, rely on filtration or absorption
to remove airborne contaminants released in the cabinet, so that the air may be
safely discharged back into the laboratory. Recirculating fume cabinets are not
normally applicable for use with radioactive materials.
Any installed fume cupboards must meet the requirements of the local
appropriate standard and any air discharged to the atmosphere must meet the
requirements of the appropriate regulatory authority. The standard will usually
specify the minimum face velocity through the working aperture (e.g. 0.5 m/s).
This should be checked on a regular basis and should be measured with the


aperture fully open and in its minimum position. At the minimum position, the
face velocity may need to be higher to retain a constant exhaust rate from the
Before initial use, and as part of a regular quality control schedule, a smoke
test should be performed. This is to provide visual evidence of fume containment
within, or escape from, the fume cupboard. Smoke is released in and around the
fume cupboard and the visual pattern of airflow is observed. The results of the
smoke test must be documented and any reverse flows from the confines of the
cupboard corrected before subsequent use.

FIG.9.7. Fume cupboard suitable for use with radioactive materials.

9.7.2. Laminar flow cabinets

Laminar flow cabinets provide a non-turbulent airstream of near constant
velocity, which has a substantially uniform flow cross-section and with a variation
in velocity of not more than 20%. Laminar flow cabinets provide product
protection while a fume cupboard is designed to provide operator protection. The
air supplied to the cabinet is usually passed through a high efficiency particulate
air filter, which is designed to remove 99.999% of particles greater than 0.3 m in
size. It must be remembered that the laminar flow of air (usually vertical) will be
disturbed by the presence of any objects within the cabinet, including shielding


and the arms of the operator. During use, the filtered air may escape from the
front of the cabinet, when the airflow is disturbed, so operator protection cannot
be ensured.
9.7.3. Isolator cabinets
Isolator cabinets provide both operator and product protection. Figure 9.8
shows an example of a blood cell labelling isolator. The product is manipulated
through glove ports so that the interior of the cabinet is maintained totally sterile
and full operator protection is provided. Airflow within the isolator is deliberately
designed to be turbulent so that there are no dead spaces within the cabinet. The
unit illustrated incorporates a centrifuge which can be controlled externally. A
dose calibrator can be included within the isolator, so that the cell suspension
does not need to be removed from the isolator for the activity to be measured.
The isolator incorporates timed interlocks on the vacuum door seals to ensure
that the product remains sterile.

FIG.9.8. Blood cell labelling isolator (courtesy of Amercare Ltd).




9.8.1. Shielding for , and positron emitters
Shielding will be required in the walls of the radiopharmacy, in any
containment enclosures, in a body shield to protect the operator at the dispensing
station, and around individual vials and syringes containing radionuclides.
Shielding of the walls of the radiopharmacy can be minimized by appropriate
local shielding around the sources being handled. Shielding may be constructed
from a variety of materials, including lead and concrete in walls, lead or tungsten
in local shielding for emitting radionuclides, and aluminium or Perspex for
pure emitters. For positron emitters, the shielding will be primarily determined
by the 511 keV annihilation photons rather than by the positrons themselves.
A low atomic number material, such as aluminium or Perspex, is used for pure
emitters since this minimizes the production of bremsstrahlung radiation. As
radiation has a finite range in materials, determined by the maximum energy,
the thickness of the shielding needs to be greater than this range to ensure that all
of the particles are absorbed. Polymethyl methacrylate (Perspex or lucite) has
a density of 1.19 g/cm3, similar to the density of tissue and water, and is highly
suitable for absorbing particles. Table 9.6 gives the maximum energy and the
range in water for four pure emitters used in nuclear medicine.
Emax (MeV)

Range in water (mm)














The highest surface dose rates encountered in the radiopharmacy are likely
to be from 99Mo/99mTc generators which may contain >100 GBq of 99Mo. The
primary emission from 99Mo has an energy of 740 keV, so requires several
centimetres of lead shielding to reduce the dose rates to an acceptable level. The
generator, as supplied, will already contain substantial shielding but additional
shielding will usually be required. This may be available from the generator


supplier specifically designed for their generator or it may be necessary to

construct or purchase an additional shield. Figure 9.9 shows a generator supplied
by ANSTO in Australia inside a dedicated lead garage. The body of the
radiochemist is shielded by the garage doors while she is attaching the shielded
elution vial prior to an elution of the generator. These shields are heavy (>20 kg),
so it is important that the bench surfaces are strong enough to support the weight.
The generator is itself quite heavy, so mechanical lifting devices may be necessary
to prevent back injuries to staff when lifting the generator into position.

FIG.9.9. Lead garage surrounding a 99Mo/99mTc generator.

Vials of radiopharmaceuticals must be kept shielded. The shields are usually

constructed so that only the rubber septum of the vial is accessible, thereby
protecting the hands of the operator during dispensing (see Fig. 9.10). The vials
themselves should never be held by the fingers once they contain radioactivity,
instead long forceps should always be used (see Fig. 9.11).



FIG.9.10. Shielded vial used to hold reconstituted radiopharmaceuticals.

FIG.9.11. Using long forceps to handle a vial containing radioactivity.



During radiopharmaceutical preparation, dispensing and administration

to the patient, the activity is usually manipulated in syringes. The dose rates at
the surface of the syringes may exceed 1 Sv s1 MBq1 depending on the
volume of liquid and the size of the syringe. The plastic of the syringe provides
little absorption of any high energy particles, and for radionuclides used for
therapy, the surface dose rates will be in excess of 10 mSv/s, so that the annual
dose limit of 500 mSv to the extremities can easily be exceeded. Syringes should
never be more than half filled, so that the syringe can be picked up near the
plunger where the fingers are not over the activity. Syringe shields must be used
whenever possible. These must be made of Perspex for the pure emitters and
of lead or tungsten for the emitters (see Fig. 9.12). A lead glass window is
necessary to permit observation of the contents of the syringe. Syringe shields
with a spring-loaded catch to hold the syringe in place are preferable to those
using a screw, as the screw-thread wears quickly with use.

FIG.9.12. A tungsten syringe shield for emitting radionuclides and a Perspex syringe shield
for pure emitters.




of lead (mm)





















6.52 104






5.09 10






3.97 106











































































4.11 104





5.00 104







3.10 10



1.71 10



9.04 10

511 keV

The transmission data for 201Tl includes a contribution of 1.5% of the contaminant 200Tl, the
maximum level likely to be encountered in clinical practice.

9.8.2. Transmission factors for lead and concrete

Section 1.6 indicates that the attenuation of monoenergetic photons through
materials such as lead or concrete will be exponential, characterized by the linear
attenuation coefficient or the half-value layer (HVL). However, this is only
correct for narrow beam geometries, using collimated beams of radiation, which


are rarely encountered in practice. Furthermore, the attenuation of the radiation

from radionuclides which emit more than one photon, such as 67Ga and 131I,
cannot be expressed as a simple HVL.
Tables 9.7 and 9.8 give the measured broad beam transmission factors for
lead and concrete for five radionuclides used in nuclear medicine and for 511 keV
photons from positron emitters. This information can be used to calculate the
required thickness of shielding around vials, for the body protection of the
operator and for the walls of the radiopharmacy. The values for 201Tl include a
contribution from 200Tl, a common contaminant, which has prominent energies at
368 keV and 1.2 MeV. A contribution of 1.5% of 200Tl has been included which is
the maximum likely value at the time of calibration.
(DENSITY: 2.35 g/cm3) (cont.)
Thickness of
concrete (mm)
































































































































511 keV




(DENSITY: 2.35 g/cm3) (cont.)
Thickness of
concrete (mm)









5.60 104





511 keV





4.61 10



3.80 106 9.16 105 7.70 104 9.39 105 8.95 104

6.30 10


4.95 10


The transmission data for 201Tl includes a contribution of 1.5% of the contaminant 200Tl, the
maximum level likely to be encountered in clinical practice.


Every radiopharmacy is unique and there is no one design that can be used
in all situations. The requirements of a single camera practice using only 99mTc
radiopharmaceuticals will be very different from a large teaching hospital with
PET facilities and in-patient radionuclide therapy rooms. However, in addition
to the general building requirements given in section 3.1.3 of Ref. [9.10], there
are some general rules specific to a radiopharmacy that can be applied in most
The radiopharmacy should be located in an area that is not accessible to
members of the public.
There should be easy access from the radiopharmacy to the injection rooms
and imaging rooms to minimize the distance that radioactive materials need
to be transported.
The radiopharmacy should not be adjacent to areas that require a low and
constant radiation background such as a counting room.
There should be an area within the radiopharmacy designated as a
non-active area that is used for record keeping and/or computer entry.
A refrigerator will be required for the storage of lyophilized
radiopharmaceutical kits. A laboratory-grade unit is preferred to ensure that
the temperature remains constant.
A dedicated dispensing area with a body shield and lead glass viewing
window will be required. This will normally be adjacent to the dose
calibrator, so that the dispensed activity can be measured while the operator
is still protected by the body shield. The thickness of the shield and window
will depend on the radionuclide or radionuclides in use. PET radionuclides



will require substantial thickness and lead glass should be supplied as a

single block rather than as a stack of thinner sheets.
A storage area will be required for reconstituted radiopharmaceuticals, in
shielded containers, together with radiopharmaceuticals purchased ready
for dispensing such as 67Ga-citrate and 201Tl-chloride.
The radiopharmacy must contain facilities for radioactive waste disposal.
This will normally include separate shielded storage bins for short lived
radionuclides such as 99mTc and for radionuclides with longer half-lives
such as 131I. In addition, there must be shielded containers for sharps, such
as syringes with needles. A separate shielded storage bin may be required
if a large number of bulky items, such as aerosol or Technegas kits, need to
be stored.
If a Mo/Tc generator is used, this should be positioned away from the
dispensing area to minimize the dose received by the person dispensing the
radiopharmaceuticals. Some countries require the generator to be housed
inside a laminar flow cabinet. All local regulatory requirements must be
taken into account when designing the radiopharmacy.
If cell labelling procedures are to be performed, a dedicated area with a
laminar flow cabinet or isolator will be required to ensure that the product
remains sterile during the labelling procedure.
A fume cupboard, together with an activated charcoal filter on the exhaust,
will be required if radioiodination procedures are to be performed.
Some radiopharmaceuticals require a heating step in their preparation.
This is often performed using a temperature controlled heating block. This
must be in a dedicated separately shielded area, particularly as several
gigabecquerels of 99mTc are often involved. Similarly, the radiolabelling of
blood samples may require local shielding of mixers and centrifuges.
Wall, floor and ceiling surfaces should be smooth, impervious and durable,
and free of externally mounted features such as pipes or ducts to facilitate
any radioactive decontamination.
Bench surfaces should be constructed of plastic laminate or resin composites
or stainless steel, and benches must be able to safely withstand the weight
of any required lead shielding.
Hand washing facilities must be available which can be operated without
the use of the operators hands to prevent the spread of any contamination.
An eye-wash should also be available.
A contamination monitor must be available in a readily accessible location.
A wall-mounted monitor to check for any hand contamination should be
mounted near the exit from the radiopharmacy. A model which can be
removed and used as a general contamination monitor is useful.




Until relatively recently, the safety of the staff when handling and storing
radioactive materials was the sole concern when designing a radiopharmacy. The
security of the radioactivity was often not specifically addressed. Unfortunately,
it is now apparent that radioactive materials can be used for malicious purposes
and the security of the radiopharmacy must now be considered.
The IAEA has categorized radioactive sources on a scale of 1 to 5, based
on activity and nuclide, where category 1 is potentially the most hazardous.
Sources categorized as 1, 2 or 3 are known as security enhanced sources. The
security measures in place for safety purposes are considered adequate to ensure
the physical security of category 4 and 5 sources. Legislation is now, or will be,
in place in each jurisdiction to address the security of security enhanced sources.
This currently only applies to sealed sources, and no sealed sources used in
nuclear medicine are categorized as either 1, 2 or 3. However, the principles can
be applied to unsealed sources. A Mo/Tc generator with an activity of greater
than 300 GBq is a category 3 source.
Radioactive materials are at most risk of being stolen or lost when they
are being transported to and from the facility. They will be in the appropriate
transport container and, therefore, can be easily handled by someone with
malicious intent. It is essential that all consignments of radioactive materials to
the nuclear medicine facility are left in a secure area and not left, for example,
on a loading dock. During working hours, all deliveries must be signed for by a
designated staff member and the material safely unpacked and stored within the
department. Some deliveries may occur outside of working hours. In this case,
a dedicated secure area must be provided where the radioactive materials can
be left. A key could be provided to the supplier for this area only, so that the
radioactivity can be safely and securely delivered, but access to other parts of the
facility is prevented. The supplier may need to be accompanied by the facilitys
security staff when delivering the shipment.
Whether secure access (such as electronic card access) to the radiopharmacy
during working hours is required will depend on local requirements and the
layout of the nuclear medicine department. It is essential that only trained nuclear
medicine staff have access to the radiopharmacy. The need for controlled access
needs to be balanced against the possibility of inadvertent contamination of the
door or access mechanism by staff returning to the radiopharmacy.




The local regulations may specify the minimum records that must be kept
at the facility, the form in which these must be kept (paper and/or electronic)
and the time for which the records must be kept. Records can be generated as
part of the quality assurance (QA) programme, for the receipt and subsequent
administration of a radiopharmaceutical to a patient, and for waste disposal.
9.11.1. Quality control records
A key element of any QA programme is proper record keeping, so that
any long term trends associated with a particular item of equipment or batch of
radiopharmaceuticals can be identified and acted on before image quality and/or
patient dose are compromised. Records should, at the very least, include details
Acceptance testing of the dose calibrator;
All constancy tests;
Radiopharmaceutical testing.
Failures identified at acceptance or constancy testing and
radiopharmaceutical testing, and the actions taken to remedy those failures,
should be documented and these records kept for the lifetime of the equipment.
The following records should be kept for all generator elutions:
Time of elution;
Volume of eluate;
Technetium-99m activity;
Molybdenum-99 activity;
Radionuclidic purity.
9.11.2. Records of receipt of radioactive materials
Complete records of the radionuclide, activity, chemical form, supplier,
suppliers batch number and purchase date should be kept. On arrival, if a
package containing radioactive material is suspected of being damaged, the
package should be:
Monitored for leakage with a wipe test;
Checked with a survey meter for unexpectedly high external radiation


If a package is damaged or suspected of being damaged, the supplier should

be contacted immediately, and the details recorded.
9.11.3. Records of radiopharmaceutical preparation and dispensing
The preparation of radiopharmaceuticals needs to be performed in
accordance with the manufacturers requirements as specified in the product
documentation, including any quality control such as thin-layer chromatography.
Records of each preparation should include the:
Name of the radiopharmaceutical;
Cold kit batch number;
Date of manufacture;
Batch number of final product;
Radiochemical purity results;
Expiry date.
A record for each patient dose dispensed must be kept with the:
Name of the patient;
Name of the radiopharmaceutical;
Measured radioactivity;
Time and date of measurement.
All unit patient doses (syringes, capsules or vials) supplied by a central
radiopharmacy should identify the patients name and the radionuclide and
radiopharmaceutical form. These should be verified on arrival and the activity
should be confirmed in a dose calibrator prior to administration to the patient,
and recorded as above.
9.11.4. Radioactive waste records
Radioactive waste generated within a nuclear medicine facility usually
consists of radionuclides with half-lives of less than one month. This waste will
normally be stored on-site, be allowed to decay to background radiation levels
and then be disposed of as normal waste or biologically contaminated waste
(see Section 3.4.7). It is, therefore, not normally necessary to keep records of
radioactive waste disposal from the facility, but it will be necessary to keep
records of the waste in storage while it decays. In some circumstances, the waste
will contain a single known radionuclide, such as 131I from patients receiving
radioiodine ablation therapy. In many cases, the waste will contain a mixture of


short lived radionuclides. Each package of waste (bag, sharps container, wheeled
bin) must be marked with the:
Radionuclide, if known;
Maximum dose rate at the surface of the container or at a fixed distance
(e.g. 1 m);
Date of storage.
The above information should be recorded, together with information
identifying the location of the container within the store, and the likely release
date (e.g. ten half-lives of the longest lived radionuclide in the container).
When the package is finally released for disposal, the record should be
updated to record the dose rate at that time, which should be at background levels,
the date of disposal, and the identification of the person authorizing its disposal.
Old sealed sources previously used for quality control or transmission
scans, such as 137Cs, 57Co, 153Gd and 68Ge, should be kept in a secure store until
the activity has decayed to a level permitted for disposal, or the source can be
disposed of by a method approved by the regulatory authority.
[9.1] TYLER, D.K., WOODS, M.J., Syringe calibration factors for the NPL Secondary
Standard Radionuclide calibrator for selected medical radionuclides, Appl. Radiat.
Isot. 59(2003) 367372.
of Ionization Chamber Systems for Assay of Radionuclides, IEC 61145:1992, IEC
Equipment Radionuclide Calibrators Particular Methods for Describing
Performance, IEC 61303:1994, IEC (1994).
Instrumentation Routine Tests Part 4: Radionuclide Calibrators, IEC/TR
61948-4:2006, IEC (2006).
[9.5] American National Standards Institute, Calibration and Usage of Dose Calibrator
Ionization Chambers for the Assay of Radionuclides, ANSI N42.13-2004, ANSI
Radiation Dose to Patients from Radiopharmaceuticals, Publication 53, Pergamon
Press, Oxford and New York (1988).




Radiation Dose to Patients from Radiopharmaceuticals (Addendum to ICRP
Publication 53), Publication 80, Pergamon Press, Oxford and New York (1998).
Radiation Dose to Patients from Radiopharmaceuticals Addendum 3 to ICRP
Publication 53, Publication 106, Elsevier (2008).
Radiological Protection and Safety in Medicine, Publication 73, Pergamon Press,
Oxford and New York (1996).
[9.10] INTERNATIONAL ATOMIC ENERGY, Applying Radiation Safety Standards in
Nuclear Medicine, Safety Reports Series No. 40, IAEA, Vienna (2005).
for Intakes of Radionuclides by Workers, Publication 30, Pergamon Press, Oxford and
New York (1979).
Recommendations of the International Commission on Radiological Protection,
Publication 60, Pergamon Press, Oxford and New York (1991).
Limits on Intake of Radionuclides by Workers Based on the 1990 Recommendations,
Publication 61, Pergamon Press, Oxford and New York (1991).

GADD, R., et al., Protocol for Establishing and Maintaining the Calibration of Medical
Radionuclide Calibrators and their Quality Control, Measurement Good Practice Guide No. 93,
National Physical Laboratory, UK (2006).
GROTH, M.J., Empirical dose rate and attenuation data for radionuclides in nuclear medicine,
Australas. Phys. Eng. Sci. Med. 19 (1996) 160167.
Limits on Radioactive Contamination on Surfaces in Laboratories, Radiation Health Series
No. 38, NHMRC (1995).
SCHRADER, H., Activity Measurements with Ionization Chambers, Monographie Bureau
International des Poids et Mesures No. 4 (1997).


Department of Medical Physics,
Memorial Sloan Kettering Cancer Center,
New York, United States of America
Historically, nuclear medicine has been largely an imaging based
specialty, employing such diverse and increasingly sophisticated modalities as
rectilinear scanning, (planar) gamma camera imaging, single photon emission
computed tomography (SPECT) and positron emission tomography (PET).
Non-imaging radiation detection, however, remains an essential component of
nuclear medicine. This chapter reviews the operating principles, performance,
applications and quality control (QC) of the various non-imaging radiation
detection and measurement devices used in nuclear medicine, including survey
meters, dose calibrators, well counters, intra-operative probes and organ uptake
probes. Related topics, including the basics of radiation detection, statistics of
nuclear counting, electronics, generic instrumentation performance parameters
and nuclear medicine imaging devices, are reviewed in depth in other chapters
of this book.
Radiation detectors encountered in nuclear medicine may generally
be characterized as either scintillation or ionization detectors (Fig. 10.1). In
scintillation detectors, visible light is produced as radiation excites atoms of a
crystal or other scintillator and is converted to an electronic signal, or pulse, and
amplified by a photomultiplier tube (PMT) and its high voltage (5001500 V).
In ionization detectors, free electrons produced as radiation ionizes a stopping
material within a sensitive volume are electrostatically collected by a bias voltage
(10500 V) to produce an electron signal. In both scintillation and ionization
detectors, the unprocessed signal is then shaped and amplified. For some types
of detector, the resulting pulses are sorted by their amplitude (or pulse height),
which is related to the X ray or ray energy absorbed in the detector.


10.2.1. Ionization detectors

Detector materials in the most common ionization detectors are gaseous
and such detectors are, therefore, often known as gas filled detectors; however,
as discussed in the following, solid state ionization detectors also exist. The
two most widely encountered gas ionization detectors in nuclear medicine are
dose calibrators and Geiger counters. The principal difference between these
detectors is the magnitude of the bias voltage between the anode and cathode, as
indicated graphically in Fig. 10.2. When the bias voltage is less than 300 V, ion
pairs (i.e. free electrons and positive ions) produced as radiation passes through
the sensitive volume may recombine, thereby preventing at least some electrons
from reaching the anode and yielding an artefactually low signal. The 0300 V
range is, therefore, called the recombination region.
Pulses in a preset pulse height range are counted
Pulse sorted by amplitude (height)
Pulse shaped and amplified

High voltage
5001500 V

Unprocessed electron signal






Bias voltage
10500 V


X or ray


X or ray
(a) Scintillation detector

(b) Ionization detector

FIG.10.1. Basic design and operating principles of (a) scintillation and (b) ionization

At a bias voltage of 300 V, all of the primary electrons (i.e. the electrons
produced directly by ionization of the detector material by the incident radiation)
are collected at the anode and the detector signal is, thereby, maximized. Since


Spontaneous discharge region

Geiger counter region


Proportional counter region


Ionization chamber region


Recombination region

Signal (total number of electrons per primary electron)


Bias voltage (V)


FIG.10.2. The signal (expressed as the amplification factor, that is, the total number of
electrons per primary electron produced in the detector material) as a function of the bias
voltage for gas filled ionization detectors. The principal difference among such detectors is the
magnitude of the bias voltage between the anode and cathode. The amplification factors and
the voltages shown are approximate.

there are no additional primary electrons to collect, increasing the bias voltage
further (up to 600 V) does not increase the signal. The 300600 V range,
where the overall signal is equivalent to the number of primary electrons and,
therefore, proportional to the energy of the incident radiation, is called the
ionization chamber region. At a bias voltage of 600900 V, however, the large
electrostatic force of attraction of the anode accelerates free electrons, as they
travel towards the anode, to sufficiently high speeds to eject additional orbital
electrons (i.e. secondary electrons) within the sensitive volume, contributing to
an increasing overall signal the higher the voltage, the more energetic the
electrons and the more secondary electrons are added to the overall signal. The
number of electrons comprising the overall signal is, thus, proportional to the


primary number of electrons and the energy of the incident radiation, and the
600900 V range is, therefore, called the proportional counter region. As the bias
voltage is increased further, beyond 900 V (up to 1200 V), free electrons (primary
and secondary) are accelerated to very high speeds and strike the anode with
sufficient energy to eject additional electrons from the anode surface itself. These
tertiary electrons are, in turn, accelerated back to the anode surface and eject even
more electrons, effectively forming an electron cloud over the anode surface and
yielding a constant overall signal even with further increase in the bias voltage.
The 9001200 V range is called the Geiger counter (or GeigerMller) region.
Importantly, the magnitude of the charge represented by this electron cloud is
independent of the number of electrons initiating its formation. Therefore,
in contrast to ionization chamber and proportional counter signals, the Geiger
counter signal is independent of the energy of the incident radiation. Finally,
beyond a bias voltage of 1200 V, atoms within the detector material are ionized
even in the absence of ionizing radiation (i.e. undergo spontaneous ionization),
producing an artefactual signal; the voltage range beyond 1200 V is known as the
spontaneous discharge region.
Although the bias voltage is the principal difference among different types
of gas filled ionization detectors, there may be other differences. The sensitive
volume, for example, may or may not be sealed. Unsealed sensitive volumes
contain only air at atmospheric (ambient) pressure. For detectors with unsealed
volumes, the signal must be corrected by calculation for the difference between
the temperature and pressure at which the detector was calibrated (usually
standard temperature and pressure: 27C and 760 mm Hg, respectively) and
the ambient conditions at the time of an actual measurement. For detectors with
sealed volumes, gases other than air (e.g. argon) may be used and the gas may be
pressurized, providing higher stopping power, and, therefore, higher sensitivity,
than detectors having a non-pressurized gas in the sensitive volume. In addition,
different geometric arrangements of the anode and cathode, such as parallel
plates (used in some ionization chambers), a wire along the axis of a cylinder
(used in Geiger counters), etc., may be used.
The functional properties and, therefore, the applications of the various
types of ionization detector ionization chambers, proportional counters and
Geiger counters are largely dictated by their respective bias voltage dependent
signal (Table 10.1). Ionization chambers are widely used in radiation therapy to
calibrate the output of therapy units and in nuclear medicine as dose calibrators
(i.e. devices used to assay radiopharmaceutical activities). The relatively
low sensitivity of ionization chambers is not a major disadvantage for such
applications, as the radiation intensities encountered are typically rather large.
The stability of the response is an important advantage, however, as it allows
the use of unconditioned AC electrical power (i.e. as provided by ordinary wall


outlets). Proportional counters, because of their need for a stable bias voltage
and, therefore, specialized power supplies, are restricted to research applications
(e.g. in radiobiology) where both higher sensitivity and the capability of energy
discrimination may be advantageous. Proportional counters often employ an
unsealed, gas flow-through sensitive volume. Geiger counters, because of their
high sensitivity and stability with respect to voltage (allowing the use of a portable
power supply such as an ordinary battery), are widely used as survey meters to
measure ambient radiation levels and to detect radioactive contamination. For
such applications, sensitivity, and not energy discrimination, is critical. As with
dose calibrators, Geiger counters have sealed sensitive volumes, avoiding the
need for temperaturepressure corrections.
In addition to the more familiar gas filled ionization detectors, solid state
ionization detectors are now available. Such detectors are based on a family of
materials known as semiconductors. The pertinent difference among (crystalline)
solids conductors, insulators and semiconductors is related to the widths of
their respective electron forbidden energy gaps. In a semiconductor, the highest
energy levels occupied by electrons are completely filled but the forbidden
gap is narrow enough (<2 eV) to allow radiative or even thermal excitation at
room temperature, thereby allowing a small number of electrons to cross the
gap and occupy energy levels among the otherwise empty upper energy levels.
Such electrons are mobile and, thus, can be collected by a bias voltage, with the
amplitude of the resulting signal being equivalent to the number of electrons
produced by the radiation and, therefore, proportional to the radiation energy.
Although many semiconductor materials have suitably large energy gaps (~2 eV),
techniques must be available to produce crystals relatively free of structural
defects. Defects (i.e. irregularities in the crystal lattice) can trap electrons
produced by radiation and, thus, reduce the total charge collected, degrading
the sensitivity and overall detector performance of semiconductors. Practical,
reasonably economical crystal growing techniques have been developed
for cadmium telluride (CdTe), cadmium zinc telluride (CZT) and mercuric
iodide (HgI2), and these detectors have been incorporated into commercial
intra-operative gamma probes and, on a limited basis, small field of view gamma







300600 V

600900 V

9001200 V

Response stable with respect to









Capable of energy




Dose calibrator


Survey meter

Bias voltage operating range



The stability with respect to the bias voltage corresponds to a constant signal over the
respective detectors operating voltage range. In contrast to ionization detectors and Geiger
counters, proportional counters are unstable with respect to the bias voltage and, thus,
require specialized, highly stable voltage sources for constancy of response.
The sensitivity of a detector is related to its amplification factor (see Fig. 10.2).
If the total number of electrons comprising the signal is proportional to the number of
electrons directly produced by the incident radiation and, therefore, proportional to its
energy, as in ionization detectors and proportional counters, radiations of different energies
can be discriminated (i.e. separated) on the basis of the signal amplitude.

10.2.2. Scintillation detectors

In scintillation detectors, radiation interacts with and deposits energy in a
scintillator, most commonly, a crystalline solid such as thallium-doped sodium
iodide (NaI(Tl)). The radiation energy thus deposited is converted to visible light.
As the light is emitted isotropically (i.e. in all directions), the inner surface of
the light-tight crystal housing is coated with a reflective material so that light
emitted towards the sides and front of the crystal are reflected back towards a
PMT (Fig. 10.3); this maximizes the amount of light collected and, therefore,
the overall sensitivity of the detector. Interposed between the back of the crystal
and the entrance window of the PMT is the light pipe, nowadays simply a thin
layer of transparent optical gel. The light pipe optically couples the crystal to the
PMT and, thus, maximizes the transmission (>90%) of the light signal from the
crystal into the PMT. When struck by light from the crystal, the photocathode
coated on the inner surface of the PMT emits electrons. Immediately beyond the
photocathode (which is at ground, that is, 0 V) is the focusing grid, maintained at
a relatively low positive voltage on the order of 10 V. As electrons pass through
the focusing grid, they are attracted by a relatively large positive voltage, ~300 V,
on the first of a series of small metallic elements called dynodes. The resulting
high speed impact of each electron results in the ejection from the dynode surface
of an average of three electrons. The electrons thus ejected are then attracted by


the even larger positive voltage, ~400 V, on the second dynode. The impact of
these electrons on the second dynode surface ejects an additional three electrons,
on average, for each incident electron. Typically, a PMT has 1012 such dynodes
(or stages), each ~100 V more positive than the preceding dynode, resulting in
an overall electron amplification factor of 310312 for the entire PMT. At the last
anode, an output signal is generated. The irregularly shaped PMT output signal
is then shaped by a preamplifier and further amplified into a logic pulse that
can be further processed electronically. The resulting electrical pulses, whose
amplitudes (or heights) are proportional to the number of electrons produced
at the PMT photocathode are, therefore, also proportional to the energy of the
incident radiation. These pulses can then be sorted according to their respective
heights by an energy discriminator (also known as a pulse height analyser) and
those pulses with a pulse height (i.e. energy) within the preset photopeak energy
window (as indicated by the pair of dashed horizontal lines overlying the pulses
in Fig. 10.3) are counted by a timer/scaler.
Advantageous features of scintillation detectors include:
High electron density (determined by mass density and effective atomic
number Zeff);
High light output;
For certain applications such as PET, speed of light emission.
High mass density and effective atomic number maximize the crystal
stopping power (i.e. linear attenuation coefficient ) and, therefore, sensitivity.
In addition, a higher atomic number crystal will have a higher proportion of
photoelectric than Compton interactions, thus facilitating energy discrimination
of photons which underwent scatter before entering the crystal. High light output
reduces statistical uncertainty (noise) in the scintillation and associated electronic
signal and, thus, improves energy resolution and scatter rejection. Other detector
considerations include:
Transparency of the crystal to its own scintillations (i.e. minimal
Matching of the index of refraction of the crystal to that of the
photodetector (specifically, the entrance window ( 1.5) of the PMT);
Matching of the scintillation wavelength to the light response of the
photodetector (the PMT photocathode, with maximum sensitivity in the
390410 nm, or blue, wavelength range);
Minimal hygroscopic behaviour.



To date, the most widely used scintillators in nuclear medicine include:

NaI(Tl), bismuth germanate (BGO), cerium-doped lutetium oxyorthosilicate
(LSO(Ce) or LSO) and cerium-doped gadolinium oxyorthosilicate (GSO(Ce) or
GSO). NaI(Tl) is used in cameras/SPECT systems, well counters and organ
uptake probes, and remains the most widely used scintillator in clinical practice;
BGO, LSO and GSO are the scintillators of choice in PET scanners because
of their higher stopping power for the 511 keV positronnegatron annihilation
photons. Thallium- and sodium-doped caesium iodide (CsI(Tl) and CsI(Na),
respectively) and cadmium tungstate as well as NaI(Tl), BGO and LSO have also
been used in intra-operative probes.
Radiation detectors may be quantitatively characterized by many different
performance parameters, particularly for those detectors such as cameras which
localize (image) as well as count radiation. For non-imaging radiation detectors
and counters, however, the most important performance parameters are sensitivity
(or efficiency), energy resolution and count rate performance (or speed).
10.3.1. Sensitivity

Sensitivity (or efficiency) is the detected count rate per unit activity (e.g. in
counts per minute per megabecquerel). As the count rate detected from a given
activity is highly dependent on the sourcedetector geometry and intervening
media, characterization of sensitivity can be ambiguous. There are two distinct
components of overall sensitivity, geometric sensitivity and intrinsic sensitivity.
Geometric sensitivity is the fraction of emitted radiations which intersect, or strike,
the detector, that is, the fraction of the total solid angle subtended at the detector
by the source. It is, therefore, directly proportional to the radiation-sensitive
detector area and, for a point source, inversely proportional to the square of the
sourcedetector distance. Intrinsic sensitivity is the fraction of radiation striking
the detector which is stopped within the detector. Intrinsic sensitivity is directly
related to the detector thickness, effective atomic number and mass density, and
decreases with increasing photon energy, since higher energy photons are more
penetrating and are more likely to pass through a detector without interacting.
Characteristic X rays and rays are emitted from radioactively decaying
atoms with well defined discrete energies. Even in the absence of scatter, however,
output pulses from absorption of these radiations will appear to originate over a
range of energies, reflecting the relatively coarse energy resolution of the detector.
For this reason, many radiation detectors employ some sort of energy-selective


counting using an energy range, or window, such that radiations are only counted
if their detected energies lie within that range (Figs. 10.3 and 10.4(a)). At least
for scintillation detectors, a so-called 20% photopeak energy window, E 10%
of E, (e.g. 126154 keV for the 140 keV ray of 99mTc) is employed, where E
is the photopeak energy of the X ray or ray being counted. For such energyselective counting, overall sensitivity appears to increase as the photopeak energy
window is widened. However, this results in acceptance of more scattered as well
as primary (i.e. unscattered) radiations.
For each radionuclide and energy window (if applicable) for which a
particular detector is used, the detector should be calibrated, that is, its sensitivity
(e.g. in cpm/MBq) S determined, at installation and periodically thereafter:

Rg R b
A 0e t


Rg is the gross (i.e. total) count rate (cpm) of the radionuclide source (RS);
Rb is the background (BG), or blank, count rate (cpm);
A0 is the activity (MBq) of the radionuclide source at calibration;
is the physical decay constant (in month1 or a1, depending on the half-life)
of the calibration radionuclide;

and t is the time interval (in months or years, respectively, again depending
on the half-life) between the calibration of the radionuclide and the current
As noted, sensitivity is highly dependent on the sourcedetector counting
geometry (including the size and shape of the source and the sourcedetector
distance), and the measured value, thus, applies exactly only for the geometry
used for the measurement.
10.3.2. Energy resolution
Energy resolution quantifies the ability of a detector to separate, or
discriminate, radiations of different energies. As illustrated in Fig. 10.4(b),
energy resolution is generally given by the width of the bell shaped photopeak,
specified as the full width at half maximum (FWHM = E) height expressed as
a percentage of the photopeak energy E, FWHM (%) =

100%. It is related




Photomultiplier tube


+ 1200 V


+ 1000 V
+ 600 V


inner surface
of crystal housing


+ 1100 V




+ 500 V



+ 400 V
+ 300 V


Energy E0


Light pipe


X or ray

FIG.10.3. The basic design and operating principle of photomultiplier tubes and scintillation

to the Poisson noise, or statistical uncertainty, inherent in the detection process.

The importance of energy resolution lies in scatter rejection, particularly for
imaging detectors. Radiation loses energy when undergoing Compton scatter
within the source and the lower energy scattered radiations may, therefore, be
discriminated from the primary radiations. However, the finite energy resolution
of radiation detectors (i.e. the width of the photopeak in the energy spectrum)
means that there will be overlap of scattered and primary radiations, as illustrated
in Fig. 10.4(a). As energy resolution improves (i.e. the FWHM (%) decreases and
the photopeak becomes narrower), the separation of unscattered and scattered
radiations increases and more counts corresponding to scattered radiation may be
eliminated, while discarding fewer counts corresponding to unscattered radiation.
10.3.3. Count rate performance (speed)
Radiation detectors have a finite dead time or pulse resolving time
typically 510 s for modern scintillation detectors and associated count
losses. The dead time is the length of time required for a counting system to


20% photopeak
energy window

Relative number of counts



Primary photons
Scattered photons
Total photons

80 100 120
Photon energy E (keV)





Maximum height

FWHM(%) =


100 = 7%



Photon energy E (keV)

E = 46 keV

Relative number of counts



662 keV



140 keV




FIG.10.4. (a) Energy spectrum for the 662 keV rays emitted by 137Cs, illustrating the
definition of energy resolution as the percentage full width at half maximum (FWHM) of the
photopeak energy E. (b) Energy spectrum for the 140 keV rays emitted by 99mTc, illustrating
the contributions of primary (unscattered) and scattered radiation counts. In (a) and (b), the
energy spectra were obtained with a thallium-doped sodium iodide (NaI(Tl)) scintillation

record an event, during which additional events cannot be recorded. As a result,

the measured count rate is lower than the actual count rate. Radiation detectors
are characterized in terms of count rate performance as either non-paralysable
or paralysable (Fig. 10.5). In non-paralysable systems, only radiation which is
actually counted prevents the counting of subsequent radiation interacting with
the detector during the dead time of that preceding radiation. In a paralysable


detector, however, even radiation which is not counted (i.e. which interacts
with the detector during the dead time of a previous event) prevents counting
of subsequent incoming radiations during the time interval corresponding to
its dead time. Geiger counters (with quenching gas) behave as non-paralysable
systems but most detectors, including scintillation detector based systems,
such as well counters, cameras and PET scanners, are paralysable. Modern
scintillation detectors generally incorporate automated algorithms to yield count
rates corrected for dead time count losses.

Observed count rate

Ideal: no dead time

count losses



True count rate

FIG.10.5. The observed versus true count rates for paralysable and non-paralysable radiation
detectors. For paralysable detectors, the observed count rate increases to a maximum value
with increasing true count rate (e.g. with increasing activity) and then decreases as the true
count rate is further increased. For non-paralysable detectors, the observed count rate also
increases with increasing true count rate, asymptotically approaching a maximum value as the
true count rate is further increased. In both cases, the maximum observed count rate is directly
related to the detectors dead time .


10.4.1. Survey meters
Survey meters, an essential component of any radiation safety programme,
are portable, battery operated, gas filled ionization detectors (or, to a much
more limited extent, solid state scintillation detectors) used to monitor ambient
radiation levels, that is, exposure rates (e.g. in coulombs per kilogram of air per
hour (C kg1 h1)) or count rates (e.g. in cpm). Among ionization detector survey
meters, so-called cutie-pies are relatively low sensitivity ionization chambers


(i.e. are operated at a relatively low potential difference between the anode
and cathode) and are designed for use where relatively high fluxes of X rays
and rays are encountered. The more familiar Geiger counters are operated at
a high potential difference (Fig. 10.2), providing a high electron amplification
factor and, thus, high sensitivity. Geiger counters are, therefore, well suited for
low level surveys, for example, checking for radioactive contamination. Both
cutie-pies and Geiger counters are generally calibrated in terms of exposure rate.
As an ionization chamber, the cutie-pies electron signal depends on the energy
of the detected X rays or rays and is, therefore, directly related to the exposure
for all radionuclides. For Geiger counters, on the other hand, signal pulses have
the same amplitude regardless of the energy of the incoming radiation. Thus,
Geiger counter calibration results apply only to the particular radionuclide(s)
used to calibrate the counter (see below). Solid state detectors employ a
non-air-equivalent crystal as the detection medium and, thus, cannot measure
exposure rates, only count rates.
10.4.2. Dose calibrator
The dose calibrator, used for assaying activities in radiopharmaceutical
vials and syringes and in other small sources (e.g. brachytherapy sources),
is a pressurized gas filled ionization chamber with a sealed sensitive volume
configured in a well-type geometry. While the intrinsic sensitivity of the dose
calibrator, as that of other gas filled detectors, is relatively low, the well-type
configuration of its sensitive volume provides high geometric efficiency1, making
the overall sensitivity entirely adequate for the relatively high radiopharmaceutical
activities (of the order of 10100 MBq) typically encountered in clinical nuclear
medicine. Dose calibrators are equipped with isotope specific push-buttons
and/or a potentiometer (with isotope-specific settings provided) to adjust for
differences in energy dependent response and to thereby yield accurate readouts
of activity (i.e. kBq or MBq) for any radioisotope.
10.4.3. Well counter
Well counters are used for high sensitivity counting of radioactive
specimens such as blood or urine samples or wipes from surveys of removable
contamination (i.e. wipe testing). Such counting results can be expressed in

The solid angle subtended at the centre of a sphere by the total surface of the sphere
is 4 steradians; a steradian is the unit of solid angle. A well-type detector configuration
approximates a point source completely surrounded by a detector, yielding a per cent geometric
efficiency of 100%, and is, therefore, referred to as a 4 counting geometry.



terms of activity (e.g. MBq) using the measured isotope specific calibration
factor (cpm/MBq) (see Eq. (10.1)). Such devices are generally comprised of a
cylindrical scintillation crystal (most commonly, NaI(Tl)) with a circular bore
(well) for the sample drilled part-way into the crystal and backed by a PMT and
its associated electronics. An alternative design for well counters is the so-called
through-hole detection system in which the hole is drilled through the entire
crystal. The through-hole design facilitates sample exchange, and because
samples are centred lengthwise in the detector, yields a more constant response
for different sample volumes as well as slightly higher sensitivity than the well
counters. In both the well and through-hole designs, the crystal is surrounded by
thick lead shielding to minimize the background due to ambient radiation.
Scintillation counters are often equipped with a multichannel analyser
for energy (i.e. isotope) selective counting and an automatic sample changer
for automated counting of multiple samples. Importantly, because of their high
intrinsic and geometric efficiencies (resulting from the use of a thick crystal and
a well-type detector configuration, respectively), well counters are extremely
sensitive and, in fact, can reliably be used only for counting activities up to
~100 kBq; at higher activities, and even with dead time corrections applied,
dead time counting losses may still become prohibitive and the measured counts
inaccurate. Modern well counters often include an integrated computer which is
used to create and manage counting protocols (i.e. to specify the isotope, energy
window, counting interval, etc.), manage sample handling, and apply background,
decay, dead time and other corrections, and, thus, yield dead time-corrected net
count rate decay corrected to the start of the current counting session.
10.4.4. Intra-operative probes
Intra-operative probes (Fig. 10.6), small hand-held counting devices,
are now widely used in the management of cancer, most commonly to more
expeditiously identify and localize sentinel lymph nodes and, thereby, reduce
the need for more extensive surgery as well as to identify and localize visually
occult disease at surgery following systemic administration of a radiolabelled
antibody or other tumour-avid radiotracer. Although intra-operative probes have
been used almost exclusively for counting X rays and rays, beta (electron and
positron) probes constructed with plastic scintillators have also been developed.
In addition, small (~10 cm) field of view intra-operative cameras have recently
become available. Intra-operative probes are available with either scintillation
or semiconductor (ionization) detectors. Scintillation detector based probes have
the advantages of relatively low cost and high sensitivity (mainly because of their
greater thickness, ~10 mm versus only ~1 mm in ionization detectors), especially
for medium to high energy photons.


Side view
Detector (crystal)
Detector thickness
Detector width

Collimator length


End view

FIG.10.6. A typical intra-operative probe (Node Seeker 900, Intra Medical Imaging LLC,
Los Angeles, CA, United States of America). (a) Hand-held detector. (b) Control and display
unit which not only displays the current count rate but also often emits an audible signal, the
tone of which is related to the count rate, somewhat analogous to the audible signal produced
by some Geiger counters. (c) A diagram of the detector and collimator assembly of a typical
intra-operative probe, illustrating that the detector (crystal) is recessed from the collimator
aperture. (Courtesy of Intra Medical Imaging LLC, Los Angeles, CA, USA.)

Disadvantages include bulkiness, and relatively poor energy resolution and

scatter rejection relative to semiconductor based probes. In some scintillation
detector intra-operative probes, the light signal from the crystal is guided to a
remote PMT through a flexible fibre-optic cable, allowing the probe assembly
to be made relatively light and compact, and more like a surgical instrument.
However, the significant loss of light in the long cable makes it more difficult to
separate scatter from unscattered X rays and rays.
On the other hand, semiconductor based probes are compact and have
excellent energy resolution and scatter rejection. To minimize structural
imperfections which degrade energy resolution, semiconductor detectors are
made relatively thin (only ~1 mm), but at the cost of lower intrinsic sensitivity.
The main disadvantage of semiconductor detectors remains their limited
thickness and resulting lower sensitivity, especially for medium to high energy
X rays and rays. Nonetheless, while scintillation detectors can be made thicker
and, therefore, more sensitive, semiconductor detectors produce more electrons
per X ray and ray stopped, and, therefore, have a superior energy resolution. To


date, the few clinical studies directly comparing scintillation and semiconductor
intra-operative probes have not provided a clear choice between the two types of
10.4.5. Organ uptake probe
Historically, organ uptake probes have been used almost exclusively for
measuring thyroid uptakes and are, thus, generally known as thyroid uptake
probes.2 Thyroid uptake (i.e. the decay-corrected per cent of administered activity
in the thyroid) may be measured following oral administration of 131I-iodide,
I-iodide or 99mTc-pertechnetate. The uptake probe is a radionuclide counting
system consisting of a wide-aperture, diverging collimator, a NaI(Tl) crystal
(typically ~5 cm thick by ~5 cm in diameter), a PMT, a preamplifier, an amplifier,
an energy discriminator (i.e. an energy window) and a gantry (stand) (Figs 10.7(a)
and (b)). Commercially available thyroid uptake probes are generally supplied as
integrated, computerized systems with automated data acquisition and processing
capabilities, yielding results directly in terms of per cent uptake.
Each determination of the thyroid uptake includes measurement of the
thyroid (i.e. neck) count rate, the thigh background count rate (measured
over the patients thigh and presumed to approximate the count contribution of
extra-thyroidal neck activity), the standard count rate (often counted in a neck
phantom simulating the thyroid/neck anatomy) and the ambient (i.e. room)
background, with a 15 min counting interval for each measurement. Based
on the foregoing measurements, and knowing the fraction of the administered
activity which is in the standard, the thyroid uptake is calculated as follows:

uptake (%) =

C neck t neck C thigh t thigh

C standard t standard C room t room

F 100% (10.2)


is the total counts;

is the measurement time;

and F is the fraction of administered activity in the standard.

At one time, organ uptake probes were also used to measure kidney timeactivity data
for the evaluation of renal function. In addition, organ uptake probes have been adapted to such
well counter applications as counting of blood samples and wipes.



FIG. 10.7. (a) A typical organ (thyroid) uptake probe system, including an integrated
computer, set-up for a thyroid uptake measurement (AtomLab 950 Thyroid Uptake System,
Biodex Medical Systems, Shirley, NY, USA). The rather large neck to collimator aperture
distance (typically of the order of 30 cm) should be noted. Although this reduces the overall
sensitivity of the measurement of the neck count rate, it serves to minimize the effect of the
exact size, shape and position of the thyroid, and the distribution of radioisotope within the
gland. (b) A diagram (side view) of the open, or flat-field, diverging collimator typically used
with thyroid uptake probes. (Courtesy of Biodex Medical Systems, Inc, Shirley, NY, USA.)

by including measurement of a standard activity with each uptake

determination, corrections for radioactive decay and day to day variation in
system sensitivity are automatic. This approach is sometimes known as the
two-capsule method, since one 131i capsule is administered to the patient while
a second, identical capsule serves as the standard and is counted with each
uptake measurement. alternatively, the patient capsule itself can be measured
immediately before it is administered and then each subsequent uptake value for
radioactive decay can be corrected from the time of measurement to the time of
administration (by multiplying the right side of eq. (10.2) by et where is the
physical decay constant of the administered isotope and t is the administration
to measurement time interval). This is sometimes known as the one-capsule
method. for either the one- or two-capsule method, the fraction of administered
activity in the standard is unity. some centres administer radioiodine as a solution,
which is more cost effective, rather than as a capsule. The standard is typically
some dilution of the administered solution and the fraction of administered
activity in the standard is, therefore, an independently determined value less than
unity; for example, if the activity in the standard solution were 1/100th of the
value in the administered solution, the value would equal 0.01.



Thyroid uptake measurements are now often performed by region of

interest analysis of planar scintigraphic images of the neck and of a standard
(i.e. phantom) acquired with a camera with parallel-hole collimation.
Organ uptake probes have also been used to measure total body activity,
for example, as part of individualized dosimetry based radioiodine treatment
of thyroid cancer. For this application, the patient may serve as his or her own
standard by measuring the patients total body count rate shortly (3060 min) after
administration of the radioisotope to allow it to disperse somewhat throughout
the body but before the patient has voided or otherwise excreted any of the
administered activity; in Eq. (10.3) below, this is designated time zero. Whole
body measurements are performed with the collimator removed from the probe,
the crystal oriented horizontally and at a height above the floor corresponding to
the mid-height of the patient, either seated or standing, at a distance of ~3 m from
the crystal. Further, anterior and posterior (i.e. conjugate-view) measurements are
performed by having the patient facing towards and then away from the crystal
for the respective measurements. The whole body activity (i.e. the per cent of
administered activity in the body) is then calculated based on the geometric mean
of the anterior and posterior count rates:
A B P B 1/2

t A t B t P t B
100% (10.3)
Total body activity (%) =


A(0) B(0) P(0) B(0)

t A(0) t B(0) t P(0) t B(0)

A and P are the anterior and posterior total body counts, respectively;
is the room (background) counts;
tA, tP and tB are the counting intervals for anterior, posterior and room counts,
and (0) indicates the same quantities at time zero.
As above, the total body activity may be corrected for radioactive decay
from the time of measurement to the time of administration (by multiplying
the right side of Eq. (10.3) by et where is the physical decay constant of the
administered isotope and t is the administration to measurement time interval).




QC, which may be defined as an established set of ongoing measurements
and analyses designed to ensure that the performance of a procedure or instrument
is within a predefined acceptable range, is a critical component of routine nuclear
medicine practice. The following is a brief review of routine QC procedures for
non-imaging nuclear medicine instrumentation.
Documenting of QC procedures and organized, retrievable records of
the results of such procedures are requirements of a sound, compliant QC
programme. A written description of all QC procedures, including the acceptable
(or tolerance) range of the results of each such procedure and the corrective
action for an out of tolerance result, should be included in the facilitys procedure
manual. For each procedure, the written description should be signed and dated
by the facility director, physicist or other responsible individual. For each QC test
performed, the following data, as well as the initials or signature of the individual
performing the test, should be recorded on a structured and suitably annotated
The test performed;
The date and time of the test;
The make, model and serial number of the device tested;
The make, model, serial number, activity at calibration and date of
calibration of any reference source(s) used;
The result(s) of the test;
A notation indicating whether the test result was or was not acceptable
(i.e. was or was not within the specified tolerance range).
Such records should be archived in chronological order in a secure but
reasonably accessible location. It is generally helpful to track the results of QC
tests longitudinally (e.g. in the form of a graph of the numerical result versus the
date of the test). In this way, long term trends in instrument performance, often
imperceptible from one day to the next, may become apparent. Increasingly, of
course, such records are maintained in electronic form. It is advisable, however,
to also maintain records in hard copy form, both as a backup and for convenient
review by regulators and other inspectors.
10.5.1. Reference sources
QC tests of nuclear medicine instrumentation are often performed not
with the radionuclides that are used clinically but with longer lived surrogate
radionuclides in the form of so-called reference sources. Such standards are


commercially available in various activities and geometries, depending on the

application. Importantly, in the USA, the certified activities of such reference
sources must be traceable to the National Institute of Standards and Technology
(NIST), formerly the National Bureau of Standards. NIST traceability helps
ensure the accuracy of the calibrated activity. As such reference sources are
long lived, a single standard may be used for months to years, avoiding the need
to prepare sources on a daily or weekly basis and, thereby, avoiding possible
inaccuracies in dispensing activity as well as the possibility of radioactive
contamination. On the other hand, as with all sealed sources, reference sources
must be checked for leakage of radioactivity (i.e. wipe-tested) periodically and
an up to date inventory of such standards must be maintained. Reference sources
are still radioactive at the end of their useful lifespan and must, therefore, be
returned to the vendor or a third party or otherwise disposed of as radioactive
A long lived radionuclide comprising a reference source must match, in
terms of the frequency and energy of its X ray and ray emissions, the clinical
radionuclide for which it acts as a surrogate in order to ensure that instrument
responses to the clinical radionuclide and to its surrogate are comparable.
Surrogate radionuclides commonly used in nuclear medicine and their physical
properties and applications are summarized in Table 10.2.
10.5.2. Survey meter
QC tests of survey meters generally include a daily battery check, with
a display indicating whether the voltage supplied by the battery is within the
acceptable operating range. In order to confirm that the survey meter has not been
contaminated (i.e. yields a reproducibly low exposure or count rate in the absence
of radioactivity), the background exposure or count rate should be measured daily
in an area remote from radioactive sources within the nuclear medicine facility,
if such an area is reasonably accessible. In addition, survey meters should be
checked daily for constancy of response by measuring the exposure or count
rate of a long lived reference source in a reproducible measurement geometry.
Aside from the short term decay of the reference source, the measured day to
day exposure or count rates should agree within 10%; if not, the meter should be
Survey meters should be calibrated that is, checked for accuracy
using suitable long lived reference sources at installation, annually and after any
repair. If the source is small (compared to the mean free path of its emitted
X rays and rays within the material comprising the source) and the distance
between the source and meter large (compared to the dimensions of the source),



then a point-source geometry is approximated and the expected dose rate D in

air is given by the inverse square law:

 = A 0e

A0 is the activity of the reference source at calibration;
is the physical decay constant of the radionuclide comprising the reference
Dt is the time interval between the calibration of the reference source and the
current measurement;
Gd is the air kerma rate constant (the subscript d indicates that only photons
with energies greater than d, typically set at 20 keV, are included) of the
radionuclide comprising the reference source;
and d is the distance between the reference source and the meter (Table 10.2).
The dose rates should be measured on each scale and, by appropriate
adjustment of the sourcemeter distance, with two readings (~20% and ~80% of
the maximum) on each scale. For all readings, the expected and measured dose
rates should agree within 10%.
Many nuclear medicine facilities have their survey meters calibrated by
the institutional radiation safety office or by a commercial calibration laboratory.
In addition to a calibration report (typically, a one page document) specifying
the reference source(s) used, the measurement procedure, and the measured
and expected exposure rates, a dated sticker summarizing the calibration results
should be affixed to the meter itself.
10.5.3. Dose calibrator
Among routine dose calibrator QC tests3, constancy must be checked
daily and accuracy and linearity at least quarterly; daily checks of accuracy
are recommended. For the constancy test, an NIST-traceable reference source,
such as 57Co, 68Ge or 137Cs (Table 10.2), is placed in the dose calibrator and the

At the installation of a dose calibrator, the geometry dependent response for 99mTc
must be measured and volume dependent (225 mL) correction factors relative to the standard
volume (e.g. 10 mL) derived. This procedure is required periodically following installation.



30 a

287 d

272 d





662 keV (86%)

511 keV (178%)

122 keV (86%)

Photopeak energy E
and frequency
of principal
X ray or ray




Air kerma rate

(mGy m2 h1 GBq1)a

Quality control
Well counter
constancy and accuracy
Dose calibrator
accuracy and constancy
Well counter
constancy and accuracy
Dose calibrator
accuracy and constancy
Well counter
constancy and accuracy
Dose calibrator
accuracy and constancy

and activity
Test tube-size rod,
~37 kBq
Vial/small bottle,
185370 MBq
Test tube-size rod,
37 kBq
Vial/small bottle,
185370 MBq
Test tube-size rod,
37 kBq
Vial/small bottle,
185370 MBq

The air kerma rate constant is equivalent to the older specific ray constant .
Germanium-68 in a sealed source is in secular equilibrium with its short lived, positron emitting daughter 68Ga (half-life: 68 min).













activity reading on each scale recorded; day to day readings should agree within
10%. for the accuracy test (sometimes also known as the energy linearity test), at
least two of the foregoing NisT-traceable reference sources are separately placed
in the dose calibrator and the activity reading on each scale recorded. for each
source, the measured activity on each scale and its current actual activity should
agree within 10%.

FIG. 10.8. Set of lead-lined plastic sleeves (CalicheckTM Dose Calibrator Linearity Test Kit,
Calicheck, Cleveland, OH, USA) for evaluation of dose calibrator linearity by the shield
method. The set is supplied with a 0.64 cm thick lead base, a colour coded unlined sleeve (to
provide an activity measurement equivalent to the zero time point measurement of the decay
method) and a six colour coded lead-lined sleeve providing attenuation factors nominally
equivalent to decay over 6, 12, 20, 30, 40 and 50 h, respectively. (Courtesy of Calicheck,
Cleveland, OH, USA.)

The quarterly check of linearity by the so-called decay method begins

with a high activity (~37 Gbq), independently calibrated 99mTc source and its
activity is assayed at 12 h intervals over three consecutive days. over that time,
equivalent to twelve half-lives of 99mTc, the activity decays to ~10 Mbq. The
measured activities are then plotted versus time on a semi-logarithmic graph and
the best fit straight line drawn through the data points thus plotted (either by
eye or using a least squares curve-fitting algorithm). for each data point, the
difference between the measured activity and the activity on the best fit straight
line at that point should be less than 10%. an alternative approach to checking
linearity is the shield method in which lead sleeves of increasing thickness
are placed in the dose calibrator with a 99mTc source (fig. 10.8). by interposing
increasing decay-equivalent thicknesses (as specified by the manufacturer for


the set of lead sleeves) between the source and the dose calibrators sensitive
volume, a decay-equivalent activity is measured for each sleeve. While the shield
method is much faster than the decay method for checking linearity (taking
minutes instead of days), an initial decay based calibration of the set of sleeves
is recommended to accurately determine the actual decay equivalence of each
10.5.4. Well counter
The routine QC tests for well counters include checks of the photopeak
energy window (i.e. energy peaking) if the counter is equipped with a
multichannel analyser, background, constancy and efficiency (or sensitivity).
Prior to counting samples containing a particular radionuclide, the energy
spectrum should be checked to verify that the counter is properly peaked,
that is, that the radionuclides photopeak coincides with the preset photopeak
energy window4. For each photopeak energy window used, the background
count rate should be checked daily. Importantly, electronic noise as well as
ambient radiation levels, which may be relatively high and variable in a nuclear
medicine facility, will produce a non-zero and potentially fluctuating background
count rate. Furthermore, even trace contamination of the counting well will
produce inaccurately high count rate values. Accordingly, a blank (i.e. an
empty counting tube or vial) should always be included to determine the current
background count. To check constancy, at least one NIST-traceable reference
source (Table 10.2) should likewise be counted each day; day to day net (i.e. gross
minus background) count rates should agree within 10%.
In addition, as noted above, for each radionuclide for which a particular
well counter is used, the counter should be calibrated that is, its efficiency
(sensitivity) (in cpm/kBq) determined at installation, annually and after any
repair (Eq. (10.1)).
10.5.5. Intra-operative probe
In addition to daily battery and background checks (as done for survey
meters), QC tests of intra-operative probes should include a daily bias check
for both the primary and any backup battery to verify that bias voltage (or
high voltage) is within the acceptable range. As intra-operative probes may not
provide a display of the energy spectrum, it may not be possible to visually check

Isotope specific radionuclide counting or imaging with a scintillation detector is

commonly performed using a 20% photopeak energy window, equivalent to an energy range of
E 10% where E is the X ray or ray energy of the radionuclide.



that the probe is properly peaked, that is, that the photopeak coincides with the
preset photopeak energy window. The lower counts or count rates resulting from
an inappropriate energy window may, therefore, go unnoticed. Thus, a long lived
reference source or set of reference sources (such as 57Co, 68Ge and/or 137Cs
(Table 10.2)) should be available for daily checks of count rate constancy; a
marked change (e.g. >10%) in the net count rate from one day to the next may
indicate an inappropriate energy window setting or some other technical problem.
Ideally, the reference sources should each be incorporated into some sort of cap
that fits reproducibly over the probe so that spurious differences in count rates
due to variations in sourcedetector geometry are avoided.
10.5.6. Organ uptake probe
Aside from differences in counting geometry and sensitivity, uptake probes
and well counters actually have very much in common and the QC procedures
checks of the photopeak energy window, background, constancy and
efficiency are, therefore, analogous. Importantly, however, efficiency should
be measured more frequently for each patient than for a well counter, so
that the probe net count rates can be reliably converted to thyroid uptakes for
individual patients.
CHERRY, S.R., SORRENSON, J.A., PHELPS, M.E., Physics in Nuclear Medicine, 3rd edn,
Saunders, Philadelphia, PA (2003).
NINKOVIC, M.M., RAICEVIC, J.J., ANDROVIC, A., Air kerma rate constants for emitters
used most often in practice, Radiat. Prot. Dosimetry 115 (2005) 247250.
ZANZONICO, P., Routine quality control of clinical nuclear medicine instrumentation: A brief
review, J. Nucl. Med. 49 (2008) 11141131.
ZANZONICO, P., HELLER, S., Physics, instrumentation, and radiation protection, Clinical
Nuclear Medicine (BIERSACK, H.J., FREEMAN, L.M., Eds), Springer Verlag, Berlin
Heidelberg (2007) 133.


Russell H. Morgan Department of Radiology and Radiological Sciences,
Johns Hopkins University,
Baltimore, Maryland, United States of America
Imaging forms an important part of nuclear medicine and a number of
different imaging devices have been developed. This chapter describes the
principles and technological characteristics of the main imaging devices used
in nuclear medicine. The two major categories are gamma camera systems and
positron emission tomography (PET) systems. The former are used to image
rays emitted by any nuclide, while the latter exploit the directional correlation
between annihilation photons emitted by positron decay. The first section of this
chapter discusses the principal components of gamma cameras and how they
are used to form 2D planar images as well as 3D tomographic images (single
photon emission computed tomography (SPECT)). The second section describes
related instrumentation that has been optimized for PET data acquisition. A major
advance in nuclear medicine was achieved with the introduction of multi-modality
imaging systems including SPECT/computed tomography (CT) and PET/CT. In
these systems, the CT images can be used to provide an anatomical context for
the functional nuclear medicine images and allow for attenuation compensation.
The third section in this chapter provides a discussion of the principles of these
11.2.1. Basic principles
The gamma camera, or Anger camera [11.1], is the traditional workhorse of
nuclear medicine imaging and its components are illustrated in Fig.11.1. Gamma
camera systems are comprised of four basic elements: the collimator, which
defines the lines of response (LORs); the radiation detector, which counts incident
photons; the computer system, which uses data from the detector to create 2D


Aluminium cover
Scintillation crystal
Exit window
Photomultiplier tube array
Shaping and positioning

Corrections and framing

FIG.11.1. Schematic diagram showing the major components of a gamma camera.

histogram images of the number of counted photons; and the gantry system, which
supports and moves the gamma camera and patient. The overall function of the
system is to provide a projection image of the radioactivity distribution in the
patient by forming an image of rays exiting the body. Forming an image means
establishing a relationship between points on the image plane and positions in the
object. This is sometimes referred to as an LOR: ideally, each position in the
image provides information about the activity on a unique line through the object.
In gamma cameras, single photons are imaged, in contrast to PET where pairs of
photons are imaged in coincidence. Thus, in order to define an LOR, a lens is
required, just as in optical imaging. However, the energies of rays are so high
Profile through image
Image plane
Point sources
FIG.11.2. The drawing on the left demonstrates the image of two point sources that would
result without the collimator. It provides very little information about the origin of the photons
and, thus, no information about the activity distribution in the patient. The drawing on the
right illustrates the role of the collimator and how it defines lines of response through the
patient. Points on the image plane are uniquely identified with a line in space.



that bending them is, for all practical purposes, impossible. Instead, collimators
are used to act as a mechanical lens. The function of a collimator is, thus, to define
LORs through the object. Figure11.2 illustrates the function of and need for the
collimator, and, thus, the basic principle of single photon imaging.
11.2.2. The Anger camera Collimators
As mentioned above, the collimator functions as a mechanical lens: it
defines LORs. The collimator accomplishes this by preventing photons emitted
along directions that do not lie along the LOR from reaching the detector. Thus,
collimators consist of a set of holes in a dense material with a high atomic
number, typically lead. The holes are parallel to the LORs. Ideally, each point in
the object would contribute to only one LOR. This requires the use of collimator
holes that are very long and narrow. However, such holes would allow very few
photons to pass through the collimator and be detected. Conversely, increasing
the diameter or decreasing the length of the holes results in a much larger range
of incident angles passing through the collimator. As illustrated in Fig.11.3, this
results in degraded resolution. As can be seen from this figure, each hole has a
cone of response and the diameter of the cone of response is proportional to the
distance from the face of the collimator.
As discussed above, changing the diameter of collimator holes changes
the resolution and also the number of photons passing through the collimator.
The noise in nuclear medicine images results from statistical variations in the
number of photons counted in a given counting interval due to the random nature
of radiation decay and interactions with the patient and camera. The noise is
described by Poisson statistics, and the coefficient of variation (per cent noise)
is inversely proportional to the square root of the number of counts. Thus,
increasing the number of counts results in less noisy images. As a result, there
is an inverse relationship between noise and spatial resolution for collimators:
improving the resolution results in increased image noise and vice versa.
Another important characteristic of collimators is the opacity of collimator
septa to incident rays. In an ideal collimator, the septa would block all incident
radiation. However, in real collimators, some fraction of the radiation passes
through or scatters in the septa and is detected. These phenomena are referred
to as septal penetration and scatter. The amount of septal penetration and scatter
depends on the energy of the incident photon, the thickness and composition of
the septa, and the aspect ratio of the collimator holes. Since gamma cameras are
used to image radionuclides with energies over a wide range, collimators are
typically available that are appropriate for several energy ranges: low energy


collimators are designed for isotopes emitting photons with energies lower than
approximately 160keV; medium energy collimators are designed for energies up
to approximately 250keV; and high energy collimators are designed for higher
energies. It should be noted that in selecting the appropriate collimator for an
isotope, it is important to consider not only the photon energies included in the
image, but also higher energy photons that may not be included in the image.
For example, in 123I there are a number of low abundance high energy photons
that can penetrate through or scatter in septa and corrupt the images. As a result,
medium energy collimators are sometimes used for 123I imaging, despite the main
photopeak at 159keV.
Point sources


FIG.11.3. Illustration of the concept of spatial resolution and how collimator hole length
and diameter affect spatial resolution. The lines from the point source through the collimator
indicate the furthest apart that two sources could be and still have photons detected at the
same point on the image plane (assumed to be the back of the collimator). Thus, sources closer
together than this would not be fully resolved (though they might be partially resolved). From
this, we see that the resolution decreases as a function of distance. It should also be noted that
the resolution improves proportionally with a reduction in the width of the collimator holes
and improves (though not proportionally) with the hole length.

For multi-hole collimators, hole shape is an additional important factor in

collimator design. The three most common hole shapes are shown in Fig.11.4.
Round holes have the advantage that the resolution is uniform in all directions
in planes parallel to the face of the collimator. However, as discussed below, the


sensitivity is relatively low because there is less open area for a given resolution
and septal thickness. Hexagonal hole collimators are the most common design for
gamma cameras using continuous crystals. They have the advantage of relatively
direction independent response functions and higher sensitivity than a round hole
collimator with the same resolution and septal thickness. Square hole collimators
are especially appropriate for detectors that have pixelated crystals. Having
squares holes that match the spacing and inter-gap distance of these detectors
results in good sensitivity with these detectors. However, the resolution varies
significantly depending on the direction, being worse by a factor of a 2 along
the diagonal direction compared to parallel to the rows of holes.

FIG.11.4. Examples of the three major hole shapes used in multi-hole collimators. They are,
from left to right: round, hexagonal and square holes. In all cases, black indicates septa and
white open space. The diameter (often called flat-to-flat for square and hexagonal holes) is d
and the septal thickness is s.

Multi-hole collimators typically have many thousands of holes. The

uniformity of the image critically depends on the holes having uniform sizes and
spacing. As a result, high quality fabrication is essential. The septa must be made
of a high density, high Z material in order to stop the incident rays. Lead is
the material of choice for most multi-hole collimators due to its relatively low
cost, low melting temperature and malleability. Lead multi-hole collimators
can be divided into two classes based on fabrication techniques: cast and foil
Fabrication of cast collimators involves the use of a mould that is filled with
molten lead to form the collimator. The mould typically consists of two plates
with holes at opposing positions that match each of the holes in the collimator. A
set of pins is placed in the holes. Lead is then poured between the plates and the
entire assembly is then carefully cooled. The plates and pins are removed leaving
behind the collimator. This technology is especially well suited to making high
and medium energy collimators as well as special purpose collimator geometries.


Foil collimators are created from thin lead foils. The foils are stamped and
then glued together to build up the collimator layer by layer. Figure11.5 shows a
schematic of how two layers are stamped and glued to form the holes. It should be
noted that in the stamping the septa that are glued must be thinner than the other
walls in order to retain uniform septal thickness and, thus, maximize sensitivity.
It is clear that precise stamping, alignment and gluing is essential to form a high
quality collimator. The septa in foil collimators can be made thinner than in cast
collimators. As a result, foil fabrication techniques are especially appropriate
for low energy collimators. Understanding the fabrication technology can help
in diagnosing problems with the collimator. For example, Fig.11.6 shows an
image with non-uniformities appearing as vertical stripes in the image of a sheet
source. This was a foil collimator and the non-uniformities apparently originated
from fabrication problems, resulting in some layers having different sensitivities
compared to other layers.

FIG.11.5. Illustration of fabrication of foil collimator by gluing two stamped lead foils. It
should be noted that the foils must be stamped so that the portions of the septa that are glued
are half the thickness of the rest of the septa. Furthermore, careful alignment is essential to
preserve the hole shapes.

FIG.11.6. Uniformity image of a defective foil collimator. The vertical stripes in the image
result from non-uniform sensitivity of the collimator due to problems in the manufacturing
process. The peppery texture is due to quantum noise and is visible because the grey level was
expanded to demonstrate the non-uniformity artefacts.



Focal point

Lines of response

Lines of response
Imaging plane

Lines of response

Lines of response
Imaging plane
Imaging plane

Imaging plane
Focal point

FIG.11.7. Illustration of the four common collimator geometries: (left to right) parallel,
converging, diverging and pinhole.

A final important characteristic of collimators is the hole geometry. There

are four common geometries in nuclear medicine: parallel, converging, diverging
and pinhole, as illustrated in Fig.11.7. Parallel holes are the most commonly used
collimators. The LORs are parallel, and there is, thus, a one to one relationship
between the size of the object and image. In converging hole collimators, the
LORs converge to a point (focal point) or line (focal line) in front of the collimator,
and there is, thus, magnification of the image. These two collimator types are
referred to as cone-beam and fan-beam collimators, respectively. These are useful
for imaging small objects, such as the heart or brain, on a large camera as they
provide an improved trade-off between spatial resolution and noise. In diverging
hole collimators, the LORs converge to a point or line behind the collimator. This
results inminification of the image. Diverging hole collimators are useful for
imaging large objects on a small camera. However, they result in a poor resolution
versus noise trade-off, and are, thus, infrequently used. Pinhole collimators
use a single hole to define the LORs. In terms of geometry, they are similar to
cone-beam collimators but with the focal point between the image plane and the
object being imaged. As a result of this, the image is inverted compared to the
object. In addition, the object can be eitherminified or magnified depending on
whether the distance from the image plane to the focal point is less than or greater
than the distance from the pinhole to the object plane. Pinhole collimators provide
an improved resolution noise trade-off when objects are close to the pinhole. They
are, thus, useful for imaging small objects such as the thyroid or small animals.
Another advantage of pinholes is that there is only a single hole, referred to as an
aperture, which determines the amount of collimator penetration and scatter. As
a result, it is possible to fabricate the aperture from high density and high atomic
number materials (e.g. tungsten, gold or depleted uranium), which can reduce


collimator penetration and scatter. This makes these collimators appropriate for
imaging radionuclides emitting high energy rays such as 131I. In addition, pinhole
collimators with changeable apertures can have different diameter pinholes. This
allows selection of resolution/sensitivity parameters relatively easily.
The collimator properties can be most completely described by the collimator
point source response function (PSRF), the noise-free image of a point source in
air with unit activity using an ideal radiation detector, as a function of position
in the object space. The shape of the collimator PSRF completely describes the
resolution properties, and when normalized to unit total volume is referred to as
the collimator point spread function (PSF).
Figure11.8 shows some sample collimator PSFs for an 131I point source
imaged with a high energy general purpose collimator and a medium energy
general purpose collimator. These PSFs are averaged over the position of the
source with respect to the collimator and, thus, do not show the hole pattern. There
are several things to note from this figure. First, using a properly designed
collimator reduces septal scatter and penetration to very low levels, while they
become significant for a collimator not designed to handle the high energies of
I. Second, the response function becomes wider as a function of distance,
demonstrating the loss of resolution as a function of increasing source to
collimator distance. Finally, there is evidence of the shape of the holes, which, in
this case, were hexagonal. The shape can be barely discerned in the shape of the
central portion of the response, which is due to photons passing through the
collimator holes. The geometry of the collimator is more evident in the septal

FIG.11.8. Sample images of the point spread function for an 131I point source at (left to right)
5, 10 and 20 cm from the face of a high energy general purpose collimator (top row) and
a medium energy general purpose collimator (bottom row). The images are displayed on a
logarithmic grey scale to emphasize the long tails of the point spread function resulting from
septal penetration and scatter. The brightness of the image has been increased to emphasize
the septal penetration and scatter artefacts.



penetration and scatter artefacts. In fact, the septal penetration is highest along
angular directions where the path through the septa is thinnest, giving rise to the
spoke-like artefacts in the directions perpendicular to the walls of the hexagonal
Another useful way to describe and understand the resolution properties of
the collimator is in terms of its frequency response. This can be described by the
collimator modulation transfer function, which is the magnitude of the Fourier
transform of the collimator PSF. Figure11.9 shows some sample profiles through
the collimator modulation transfer function. It should be noted that the collimator
response does not pass high frequencies very well and, for some frequencies,
the response is zero. This attenuation of high frequencies results in a loss of
fine detail (i.e. spatial resolution) in the images. Finally, the cut-off frequency
decreases with distance from the collimator and different collimator designs have
different frequency responses.

FIG.11.9. Sample profile through the geometric modulation transfer functions (MTFs) for
low, medium and high energy (HE, ME and LE, respectively) general purpose (GP) and high
resolution (HR) collimators for a source 5 cm (left) and 20 cm (right) from the face of the

It is often desirable to summarize the collimator resolution in terms of one

or two numbers rather than the entire response function. This is often done in
terms of the width of the collimator PSRF at a certain fraction of its maximum
value. For example, Fig.11.10 shows a sample profile through a collimator
PSF and the position of the full width at half maximum (FWHM) and the full
width at tenth maximum (FWTM). To good approximation, the FWHM of a
collimator is proportional to the distance from the face of the collimator. This
holds for all distances except those very close to the collimator face, as illustrated
in Fig.11.11. The FWTM is useful for assessing the amount of septal penetration
and scatter that are present. For a Gaussian response function, which is a good



Intensity (arbitrary units)


Half maximum






Tenth maximum

-0.5 0
Distance (cm)


FIG.11.10. Plot of the total collimatordetector point spread function for a medium energy
general purpose collimator imaging 131I, indicating the positions of the full width at half
maximum (FWHM) and the full width at tenth maximum (FWTM).


FWHM (cm)








Distance from face of collimator (cm)

FIG.11.11. Plot of the full width at half maximum (FWHM) of the geometric collimator
detector point source response function including a Gaussian intrinsic response with an
FWHM of 4mm as a function of distance from the face of the collimator for the same set of
collimators described in Fig.11.9.



approximation for the combination of a Gaussian intrinsic detector response

with the geometric collimator response, the relationship between the FWHM and
FWTM is given by FWTM/FWHM 1.86. Thus, if the FWTM is substantially
larger than this factor times the FWHM, the response has been affected by factors
other than the geometric response, such as septal penetration and scatter. It should
be noted that the effects of septal penetration and scatter on the FWTM are less
visible in a PSRF than they are in a line source response function.
The FWHM of the collimator resolution can be estimated from geometric
arguments. Figure11.12 shows a schematic that can be used to derive the
resolution for a point source a distance Z from the face of the collimator. The
collimator hole has a length L and a width d. The image plane (often taken to be
the mean path of the primary photons in the crystal plus any physical gap) is a
distance B behind the back face of the collimator. The photon passing through the
collimator holes with the most oblique angle of incidence will have an incident
angle defined by tan q=d/L. Thus, the extreme limits of the response function
will be defined by this limit. If it is assumed that the geometric response function
is triangular in shape, then the FWHM in Fig.11.12 will be half of this distance.
Using similar triangles, it can be shown that the FWHM is given by:

(Z + L + B) (11.1)

Thus, it can be seen that, as described above, the FWHM is linearly related
to the distance from the face of the collimator and is proportional to that distance
when the distance is large compared to L + B.
The resolution of the collimatordetector system depends on both the
resolution of the collimator and the intrinsic resolution of the gamma camera. For
continuous-crystal gamma cameras, the intrinsic resolution can be modelled with
a Gaussian function. In this case, the total response function for the collimator
detector is the convolution of the intrinsic resolution and the collimator response.
If the collimator response is approximated by a Gaussian function, the FWHM is
given by the Pythagorean sum of the intrinsic and collimator FWHMs:
FWHM total = FWHM collimator
+ FWHM intrinsic

Figure11.11 shows a plot of the total geometric FWHM resolution

for several collimators including the effects of a 4mm Gaussian intrinsic
resolution. The curves through the points represent a fit with a function of the
above Pythagorean sum. It should be noted that except for distances close to
the collimator, the resolution is linear with distance, indicating that the total
resolution is dominated by the collimator resolution.




Image plane

FIG.11.12. Diagram illustrating the collimator geometry used to derive the expression for the
full width at half maximum.

The integral of the collimator PSRF gives the sensitivity of the collimator.
This is, in principle, a dimensionless quantity which gives the fraction of emitted
photons that pass through the collimator, and is of the order of 103104 for
typical nuclear medicine collimators. It is often useful to express the sensitivity
in terms of counts per unit activity per unit time, for example counts per second
per megabecquerel. For a parallel-hole collimator, the sensitivity is a function
of two terms: the solid angle of the hole, which is a function of (d/L)2, and the
fraction of the active area of the collimator that is open area (hole) as compared
to septa. The second term can be described in terms of the unit cell: the smallest
geometric region that can be used to form the entire collimator by a set of simple
translations. The sensitivity S of a parallel-hole collimator is given by:

aopen aopen
4 L2 a total


aopen is the open area of a collimator unit cell, i.e. the area of the hole itself;
and atotal is the total area of the cell including the part of the collimator septa lying
in the unit cell.


From the above it can be seen that, for a parallel-hole collimator, the
sensitivity is independent of the distance to the collimator face. This is because
there is a balance between the decrease in sensitivity from a single hole and the
increase in the number of holes through which photons can pass as a function
of increasing distance from the collimator face. It should also be noted that
aopen is proportional to d2, which is proportional to the FWHM resolution. The
term aopen/atotal varies slowly as a function of d if d s. Thus, the sensitivity is
proportional to the square of the resolution:
S = k FWHM 2 (11.4)

where the constant k depends only weakly on the FWHM. Since noise is directly
related to the number of counts, there is a fundamental trade-off between
resolution and noise. From the above, it is also evident that maximizing the ratio
of aopen/atotal is important in terms of reducing noise for a given resolution. Scintillation crystals
The scintillation crystal in the gamma camera converts ray photons
incident on the crystal into a number of visible light photons. The characteristics
and principle of scintillation radiation sensors are described in more detail in
Chapter6. Ideally, the crystal would be dense and of a high Z material in order
to stop all incoming rays with photoelectric events. It should have high light
output to provide low quantum noise for energy and position estimation. The
decay time of the light output needs to be fast enough to avoid a pile-up of
pulses at the count rates experienced in nuclear medicine imaging procedures.
The wavelength spectrum of the scintillation photons should be matched to the
sensitivity of the photodetectors used to convert the scintillation signal into an
electrical signal. In addition to these technical properties that directly affect
image quality, there are a number of desirable material properties that influence
the cost of the device. These include the cost of the raw material, the ease of
growing large single crystals and the sensitivity to environmental factors such as
humidity. Owing to its unique combination of desirable properties, the crystals
used in gamma cameras based on photomultiplier tubes (PMTs) are typically
made of NaI(Tl). Gamma cameras based on solid state photodetectors require
a different light spectrum and typically CsI(Tl) is used in these devices. The
scintillation properties of these materials are discussed in detail in Chapter6.
As will be described below, the interaction position of the ray with the
detector is estimated based on the distribution of the scintillation light to an
array of PMTs. It is important that the distribution of light be as independent as
possible of the depth of interaction in the crystal and depends in a predictable


way on the lateral position. Further, absorption of scintillation photons by

defects in the crystal is highly undesirable, as it will adversely affect the
accuracy and precision of energy and position estimation. Thus, in order to make
the lateral light distribution predictable, as even as possible and tominimize
absorption, Anger cameras employ a large single crystal equal to the size of the
field of view (FOV) of the camera. This can be as large as 60cm40cm in
modern cameras. As NaI(Tl) is hygroscopic, it is important to hermetically seal
the crystal in an enclosure. The back of the crystal must be optically coupled
to the photodetector, so that the back part of the crystal enclosure consists of
an optical glass exit window optically coupled to the crystal. The exit window
lies between the crystal and the photodetector array. It serves several functions.
First, it serves to hermetically seal the crystal. Second, the exit window allows
scintillation photons to pass from the crystal into the photodetector array, and,
thus, must be transparent in the emission range of the scintillator. To reduce
internal reflections, it is desirable that the index of refraction be matched
as closely as possible to that of the scintillator (n=1.85 for NaI(Tl)) and the
photodetector (n 1.5 for borosilicate glasses used in the entrance windows
of PMTs). The remainder of the enclosure should be light-tight to block out
ambient light. The front face should be thin and of a low Z material typically
Al is used to reduce the probability of incident ray absorption. Finally, to
help collect incident light photons and improve energy resolution, the inside of
the enclosure is a reflective layer. The use of specular versus diffusive reflectors
affects the nature of the light response and has an impact on the variation in the
light response on the photodetectors as a function of interaction depth, and, thus,
impacts the precision of the position estimation.
One important parameter of the scintillation crystal related to camera
performance is its thickness. The thickness is a trade-off between two
characteristics: intrinsic resolution and sensitivity. The intrinsic resolution
depends on the crystal thickness via the variation in the light distribution as a
function of depth of interaction. Since the depth of interaction can vary over
a wider range in a thicker crystal, there will be a larger variation in the light
distribution and, thus, a larger uncertainty in the estimated lateral position of
the interaction. In other words, thicker crystals generally have poorer intrinsic
resolution. The functional relationship between the thickness and intrinsic
resolution is complicated and depends on the details of the surface treatment of
the scintillator, the photodetector array and the position estimation algorithm.
For GE Millenium VG cameras, the FWHM intrinsic resolution for 140keV
photons using 0.953, 1.587 and 2.54cm thick crystals is 3.5, 3.9 and 5.2mm,



The intrinsic sensitivity of the crystal is related to crystal thickness by:

S i = 1 e t (11.5)


is the intrinsic sensitivity;

is the linear attenuation coefficient of the crystal;

and t is the crystal thickness.

Since the linear attenuation coefficient decreases with energy, the intrinsic
sensitivity also decreases with energy. Figure11.13 shows a plot of the intrinsic
sensitivity as a function of energy for several crystal thicknesses. For 140keV, the
sensitivity is ~92% for a 0.953cm (3/8 in) thick crystal. This is the most common
crystal thickness in commercial systems, though cameras with 5/8 and 1 in are
available. These crystal thicknesses provide substantially improved sensitivity
for radionuclides emitting medium and high energy photons such as 111In and 131I
at the cost of a relativelyminor reduction in intrinsic spatial resolution.

Intrinsic sensitivity

Crystal thickness

FIG.11.13. Plot of the intrinsic sensitivity of a NaI scintillation crystal as a function of energy
for several crystal thicknesses.



A final important property of the scintillation crystal is the light output. This
is a characteristic of the scintillator material, and is the number of scintillation
photons per unit energy deposited in the crystal by a photon. Thus, the total light
is proportional to the energy deposited in the crystal, and can, therefore, be used
to estimate the energy of the ray. The number of scintillation photons produced
for a given event is a Poisson random variable. Thus, the larger the number of
scintillation photons the smaller the coefficient of variation (standard deviation
divided by the mean) of the mean number of photons, and, hence, the estimated
photon energy. Thus, scintillators with high light output will provide higher
energy resolution. In addition, as will be seen below, the light distribution over
the photodetector array is used to estimate the interaction position. Since the light
collected by each element in the array is also a Poisson random variable that is
proportional to the light output, a larger light output will result in higher precision
in the estimated position, and, thus, improved intrinsic spatial resolution. One
reason that NaI(Tl) is used in gamma cameras is its high light output. Photodetector array
The next element in the radiation detector is the photodetector array.
This array measures the distribution of scintillation photons incident on the
array and converts it into a set of pulses whose charge is proportional to the
number of scintillation photons incident on each corresponding element in
the array. As described below, the output of this array is used to compute the
interaction position of the ray in the scintillator. In clinical gamma cameras, the
photodetector array is comprised of a set of 3090 PMTs arranged in a hexagonal
close packed arrangement, as illustrated in Fig.11.14. More details on the
operation and characteristics of PMTs are provided in Chapter6. In brief, PMTs
have the advantage that they are very well understood, have a moderate cost, are
relatively sensitive to low levels of scintillation light and have a very high gain. In
some commercial designs, PMTs have been replaced by semiconductor detectors
such as photodiodes. Generally, these devices are somewhat less sensitive and
have a lower gain than PMTs, resulting in more noise in the charge signal and,
thus, less precision in the energy and position estimated from the charge signal.
Since the position and energy are estimated from the set of charge signals
from the elements in the photodetector array, it is highly desirable that the
proportionality constants relating light intensity to charge be the same for all
of the photodetectors. This can be ensured by choosing matching devices and
by carefully controlling and matching the electronic gain. For PMTs, the gain is
controlled by the bias voltage applied to the tubes. Since gain is also a function of
temperature, the temperature of the photodetectors must be carefully controlled.
The gains of PMTs are very sensitive to magnetic fields, even those as small


as the Earths magnetic field. Thus, the PMTs must be magnetically shielded
using mu-metal. Finally, since the gains of tubes can drift over time, periodic
recalibration is necessary.
One of the major advantages of the gamma camera is that the number of
PMTs is much smaller than the number of pixels in images from the gamma
camera. In other words, in contrast to semiconductor detectors where a separate
set of electronics is required for each pixel, the gamma camera achieves a great
reduction in cost and complexity by estimating the interaction position of the
ray based on the output of the array of PMTs.
To understand the position estimation process, Fig.11.15 is considered.
This figure shows a cross-section through two PMTs, and the crystal and exit
window. The number of photons collected by a PMT directly (i.e. without
reflection) will be proportional to the solid angle subtended by it at the interaction
point. As can be seen in the figure, the interaction position is offset to the right,
and there is a smaller solid angle subtended by PMT 1 than by PMT 2. Thus, the
signal from PMT 1 will be smaller than for PMT 2. If the interaction position
moves to the left, so that it lies along the line separating the two PMTs, there will
be an equal amount of light collected by each PMT. The relationship between the
light collected by the two PMTs and the lateral interaction position can be used
to estimate the interaction position, as will be described in more detail below. In
addition, the total scintillation light collected by all of the PMTs is proportional
to the energy deposited by the ray in the crystal. Thus, the total charge can be
used to estimate the energy of the photon.

FIG.11.14. Cross-section of a gamma camera at the back face of the entrance window
showing the hexagonal close packed array of photomultiplier tubes. The dotted line indicates
the approximate region where useful images can be obtained.



Exit window
Interaction position

FIG.11.15. Cross-section through two photomultiplier tubes (PMTs), the exit window and
crystal in a gamma camera. The interaction position of a ray photon is indicated. The solid
angles subtended by PMT 1 and 2 are 1 and 2, respectively. Amplifiers and pulse shaping

The charge pulse from each PMT is very small and, thus, subject to
noise. In addition, the scintillation photons are emitted randomly over a finite
time (given by the scintillators decay constant), making the charge pulse rather
noisy. To make subsequent analysis of the pulse easier and more resistant to
electrical noise, the pulse is amplified and shaped prior to processing to estimate
the interaction position and photon energy. The components of this stage are a
preamplifier and shaping amplifier.
The preamplifier integrates the charge pulse from the PMT to form a
voltage pulse with height proportional to the input charge. The design of the
preamplifier should be such that the voltage height is as independent as possible
of the details of the charge pulse, such as decay and rise times. Preamplifiers are
typically mounted directly to the PMT outputs in order to avoid corruption of the
tiny charge pulses by electrical noise and interference.
Ideally, output pulses would have a very flat top to allow easy digitization
of the pulse height and be very narrow to allow high pulse rates without pulse
pile-up. However, the output pulse from the preamplifier typically has a relatively
long decay time and is not very suitable for digitization and handling high pulse
rates. As a result, the output of the preamplifier is fed into a shaping amplifier.
Typically, shaping amplifiers use a combination of integration and differentiation
stages to produce near Gaussian pulses. It should be noted that more recent
commercial gamma cameras have used digital pulse processing methods to


perform this function. This involves digitizing the output waveform from the
preamplifier. This has a number of advantages including providing the ability to
change the trade-off between energy resolution and count rate, depending on the
requirements of the particular imaging procedure. In addition, this method also
provides digital estimates of the pulse heights that can be used in digital position
and energy estimation algorithms. Position and energy estimation
The goal of the radiation detector is to provide an estimate of the energy
and interaction position of each ray incident on the detector. The output of
the photodetector array and amplifier system is a set of voltage signals for each
photon. The sum of these voltages is proportional to the gamma camera energy
and the position is a function of the set of voltage values. The position and energy
estimation circuits estimate the ray energy and position from the set of voltage
values from the photodetector array.
One way of doing this is to use a resistive network to divide the signals
from the array elements among a set of four signals often referred to as X+, X, Y+
and Y, as illustrated in Fig.11.16. The resistor values for each PMT are chosen
so that the charge is divided in proportion to its position with respect to the centre
of the array. For example, for a PMT in the centre horizontally, the resistances
for X+ and X would be equal. Similarly, for a PMT in the centre vertically, the
resistances for Y+ and Y would be equal.
Using the scheme described above, the energy E can be computed using:
E = X + + X + Y+ + Y (11.6)

However, one limitation of this method is that the total amount of light
collected is dependent on position. For example, if the interaction is directly
under a PMT, a larger fraction of the total light will be collected, resulting in a
larger value of E than if the interaction is in the gap between PMTs. This means
that the estimate of the energy will vary spatially. As discussed below, this has an
impact on camera uniformity. As a result, the energy must be corrected based on
the interaction position.
Under the assumption that the light collected by a PMT is proportional to
the distance from its centre, and with the correct resistor values, the interaction
position, defined by x and y can be computed using:


X+ X

and y =

Y+ Y


In early gamma cameras, the computations above were performed using

analogue circuits. However, in recent cameras, the pulse heights (or the entire
pulse) are digitized and the computations performed digitally.
Using the resistive summing and simple estimation approaches above
results in a number of problems. First, light collected by phototubes is not linearly
related to the distance from the interaction point. For example, the amount of light
changes relatively little for PMTs at a large lateral distance from the interaction
position. As a result, thresholding circuits are often added to exclude the signal
from PMTs with small outputs (and, thus, far from the interaction point) from
the position calculation. In addition, the distribution of light between two tubes
changes more quickly when the interaction position lies between two tubes than
it does when the interaction position is directly over a tube. This results in spatial
non-linearities where images of line sources are bent towards the centre of PMTs.
A final difficulty is that it is not possible to reliably estimate the position of
photons interacting near the edge of the camera. In this case, almost all of the
light will be collected by the nearest PMT and there will be little change in the
relative amount of light as the interaction position moves.

FIG.11.16. Illustration of a resistive network used to implement position estimation. The

output from each photomultiplier tube/preamplifier is divided by a resistive network with four
outputs, X+, X, Y+ and Y. The corresponding signals from all of the photomultiplier tubes are
connected to provide the summed signal.



As a result of the above difficulties, modern cameras use sophisticated

correction and position estimation methods. The correction methods will be
discussed in more detail below. Advanced position estimation methods involve
digitizing the outputs of all of the PMTs and using them in the position estimation.
In this case, position and energy estimation, and the various corrections are
done by a digital signal processor or microprocessor, allowing a great deal more
sophistication in the choice of algorithms. In some systems, these are done using
statistical estimation techniques such as maximum-likelihood estimation.
As alluded to above, the values of the interaction position (x, y) and energy
E are computed using equations similar to those shown above. The input variables
X+, X, Y+ and Y are related to the charge signal from the PMT. These signals
are proportional to the number of photoelectrons emitted from the photocathode.
The emission of photoelectrons is the end result of a series of random processes
that includes the number of scintillation photons produced, the number of these
collected by the PMT, the number of photoelectrons produced for a given photon
and the number of photoelectrons emitted which are focused onto the first
dynode. The net result is that there is statistical variation in the values of the
input variables for a given interaction position and energy. Thus, the values of
position and energy are not exact, and are only estimates of the true quantities.
Thus, there will be imprecision in the energy and position estimates resulting in
finite intrinsic energy and spatial resolutions (seeChapter8 for more details).
To a good approximation, both the energy and intrinsic spatial resolution
can be characterized by a Gaussian function. The energy resolution results from
variations in the total number of photoelectrons incident on the first dynode. The
random variations can be approximated by a Poisson distribution and the variance
in the energy resolution is, thus, approximately equal to the number of these
photoelectrons. The approximate energy resolution of a gamma camera can, thus,
be estimated as follows. A NaI(Tl) crystal produces, on average, 38 photons/keV.
The quantum efficiency (fraction of incident scintillation photons that produce
photoelectrons) of a PMT for the 415 nm emission of NaI(Tl) is approximately
12%. Thus, for a 140keV photon, the number of photoelectrons collected
is 140380.12=638 electrons. The FWHM is equal to approximately
2.35 times the standard deviation, so the fractional energy resolution is equal to
2.35 638 / 638 = 9.3%. This is approximately equal to the energy resolution for
a state of the art scintillation camera. It should also be noted from the above that
the energy resolution is proportional to E0.5 and spatial variations in the collection
efficiency will produce spatial variations in the energy resolution. Estimating the
intrinsic spatial resolution is more difficult than estimating the energy resolution
because of the more complicated estimation equation. However, typical intrinsic
spatial resolutions are in the range of 35mm, depending on the number of PMTs
used and details of the estimation procedure.

As mentioned above, the energy and position estimation are non-ideal,
resulting in errors in energy and position estimates. These errors give rise to
non-uniform sensitivity in the camera. Thus, to obtain clinically acceptable
images, energy, spatial and uniformity corrections are needed. The need for
these corrections is illustrated in Fig.11.17. An image is shown resulting from
a uniform distribution of rays on the camera with the collimator removed,
often referred to as an intrinsic flood image. The substantial non-uniformity, the
presence of edge packing artefacts near the edge of the FOV, and the visibility of
the tube pattern should be noted.
Edge packing artefact

Tube pattern

FIG.11.17. Intrinsic flood image of gamma camera without energy, spatial or sensitivity

Energy corrections are needed because the estimated energy depends on

spatial position. This behaviour can be understood in terms of variations of the
fraction of the scintillation light collected as a function of interaction position.
Since the energy is proportional to the light collected, differences in the fraction
of light collected will result in a proportional change in the estimated energy.
Figure11.18 shows an example energy spectrum where the fraction of collected
light was 2% lower or 2% higher than the average. This results in energy spectra


shifted to lower or higher energy, respectively. Only photons falling within

the acquisition energy window, in this case having a full width of 20% of the
photopeak energy, centred at 140keV, are counted. There are about 1.7% fewer
photons counted in the two sample energy spectra than for the case of the average
energy spectrum. Thus, the sensitivity is about 2% lower at these points. A
typical energy correction algorithm measures the energy spectrum as a function
of position in the image using a source or sources with known energies. A linear
or higher order correction is then made to the estimated energy.
Spatial corrections are needed because of biases in estimated interaction
positions. These biases result, as described above, from discrepancies in the
formulas used to estimate the position and the actual behaviour. As mentioned,
these usually result in lines being bent towards space between PMTs. Typically,
separate corrections are made for the axial (y) and transaxial (x) directions. These
corrections involve imaging a mask with a grid of holes or lines in combination
with a flood source. This results in a series of bright spots or lines in the image
that correspond to the known positions of the holes or lines in the mask. A
function, typically a polynomial, is fit to the set of true points as a function of
the set of measured points. This function can then be used to correct a measured



2% lower
2% higher

Energy (keV)



Fig.11.18. Sample energy spectrum for 140keV photons for the cases of average, 2% lower
than average and 2% higher than average light collection efficiency. The variation in light
collection efficiency results in a shift of the energy spectrum, which results in non-uniform



The final type of correction applied is a uniformity or sensitivity correction.

The goal of this correction is to make images of a flood source as uniform
as possible (seeFig.11.19). There are two types of uniformity corrections:
intrinsic, which corrects only for non-uniform sensitivity of the detector system
(i.e. excluding the collimator), and extrinsic, which corrects for both detector and
collimator non-uniformities. Uniformity corrections are made using a high-count
flood image. Uniformity correction is implemented by, in essence, multiplying
each pixel in acquired images by a factor equal to the average counts in the active
portion of the flood image divided by the counts in the corresponding pixel in the
flood image.
The number of counts in the flood image is critical in determining the
ultimate uniformity of the image. This is especially important in SPECT where
local non-uniformities can result in ring artefacts. To achieve this, the counts in
the flood image should be such that the relative standard deviation (coefficient
of variation) of the pixel counts resulting from Poisson counting statistics is less
than the desired uniformity. For example, if uniformity correction to better than
1% is desired, the average number of counts per pixel in the uniformity flood
should be greater than 1/0.012=10 kcounts per pixel. The total number of counts
in the flood, thus, depends on the number of active pixels in the image. Since
non-uniformities are generally relatively low frequency, this restriction can be
relaxed to some degree by the use of low pass filtering applied to the flood image.
Intrinsic flood images are usually acquired using a point (or syringe) source
containing a small quantity of the isotope of interest. If the source is placed at
a distance of more than five times the largest linear dimension of the camera

FIG.11.19. Intrinsic flood images for a gamma camera having a poor (left) and good
(right) set of corrections applied. It should be noted that the images are windowed so that
the brightness represents a relatively small range of count values in order to amplify the
differences. The peppery texture is due to quantum noise and is to be expected. The quantum
noise is exaggerated because of the windowing used, and can be reduced by acquiring very
high-count flood images.



FOV from the camera face, then the irradiation of the camera can be considered
uniform. Since the uniformity of the camera will, in general, vary depending
on the energy of the isotope and energy window used, this correction should
ideally be made for each isotope and energy window used. The count rate for the
acquisition should be within acceptable limits to avoid high count rate effects.
Extrinsic flood images are made using a flood or sheet source. Fillable
flood sources have the advantage that they can be used for any isotope. However,
great care must be made in filling the phantom to remove bubbles, mix the
activity and maintain a constant source thickness. In addition, images must be
obtained for each collimator used with a given isotope. As a result, 57Co sheet
sources are often used to obtain extrinsic flood images. These have the advantage
of convenience but are, strictly speaking, valid for only a single isotope.
One way to take advantage of both approaches is to perform uniformity
corrections using a combination of intrinsic flood images for each isotope
used and an extrinsic flood image obtained for the collimator in question. This
approach assumes that collimator uniformity is independent of energy and can,
thus, be measured with, for example, a 57Co sheet source. Not all equipment
vendors support this approach. Some vendors assume that the energy and linearity
corrections produce uniformity that is energy independent, and they, thus,
recommend only the use of an extrinsic flood image for uniformity correction.
Another approach is to first confirm the uniformity of all collimators via a sheet
source flood image with intrinsic correction for 57Co. Then, an extrinsic flood
image for each isotope used is acquired and used in uniformity correction for that
isotope, assuming that the collimator is sufficiently uniform. The best approach
depends on the characteristics of the individual camera. Image framing
The final step in generating gamma camera images is image framing.
Image framing refers to building spatial histograms of the counts as a function of
position and possibly other variables. This involves several steps, and is typically
done either by microprocessors in the camera or in an acquisition computer.
In this step, position is mapped to the elements in a 2D matrix of pixels. The
relationship between pixel index and physical position is linearly related to the
ratio of the maximum dimension of the FOV of the camera to the number of
pixels, the zoom factor and an image offset. The zoom factor allows enlarging
the image so that an object of interest fills the image. This can be useful, for
example, when imaging small objects. It results in a pixel size in the image that
is a factor of 1/zoom factor as large as in the unzoomed (zoom factor equals
one) image. It should be noted that even though the pixel size is decreased, the
resolution of the image will not necessarily be improved as long as the original


pixel size is smaller than the intrinsic resolution. For example, if the native pixel
size is 3.2 and the intrinsic resolution is 4mm, a zoom factor of two will result in
a pixel size of 1.6mm, but the intrinsic resolution will still be 4mm. An image
offset can be used to shift the image, so that an object of interest is in the centre
of the acquired image.
In addition to adding counts to the appropriate pixel spatially, the framing
algorithm performs a number of other important functions. The first is to reject
photons that lie outside of the energy window of interest. This is done to reject
scattered photons. Gamma cameras typically offer the ability to simultaneously
frame images corresponding to more than one energy window. This can be useful
for isotopes having multiple photopeaks, for acquiring data in energy windows
used by scatter compensation algorithms or for acquiring simultaneous images
of two or more radionuclides. Framing software typically enables the summation
of photons from multiple, discontiguous energy windows into one image as well
as simultaneously framing multiple images from different energy windows into
different images. There is often a limited number of energy windows that can
be framed into a single image, and a limit on the number of images that can be
framed at one time. These limits may depend on the image size, especially if the
framing is done by a microprocessor in the camera that has limited memory.
A second important function provided by the framing system is the ability to
obtain a sequence of dynamic images. This means that photons are recorded into
a set of images depending on the time after the start of acquisition. For example,
images could be acquired in a set of images with a frame duration of 10 s. Thus,
for the first 10 s, photons are recorded into the first image; for the second 10 s,
they are recorded into a second image; and so on. Thus, just as multiple images
are obtained in the case of a multi-energy window acquisition, multiple images
are obtained corresponding to a sequential set of time intervals. This is illustrated
in Fig.11.20, where seven dynamic frames are acquired for a time interval T.
Dynamic framing is used for monitoring processes such as kidney function,
gastric emptying, etc. The time frames are often not equal in duration as there
may be more rapid uptake at early times and a later washout phase in which the
change in activity with time is slower. The number or acquisition rate of dynamic
frames is often limited due to constraints in framing memory and this limit can
depend on the image size.
Gated acquisition is similar to dynamic acquisition in that photons are
recorded into a set of images depending on the time they are detected. However,
in gated acquisition, the time is relative to a physiological trigger, such as an
electrocardiogram (ECG) signal that provides a signal at the beginning of each
cardiac cycle. This is appropriate for processes that are periodic. The photons
are counted into a set of frames, each of which corresponds to a subinterval of
the time between the two triggers. For example, the bottom two illustrations


in Fig.11.20 should be considered. In both cases, the interval between gates is

divided into four subintervals (in cardiac imaging, 8 or 16 frames are typically
used, but 4 are illustrated in this example for simplicity). The photons arriving
in each of the subintervals are counted in the corresponding frame. Thus, for
cardiac imaging, the activity distribution during the first quarter of the cardiac
cycle is imaged in frame 1, the second quarter in frame 2, etc. This is useful for
assessing wall motion and thickening. Just as in dynamic and multiple window
acquisitions, there may be limits on the size and number of frames due to limits
in framing memory on the acquisition computer.
Since the gate is derived from a physiological signal, and physiological
signals are not necessarily exactly periodic, the framing algorithm must handle
the case of variable time intervals between gate signals. In cardiac imaging,
this corresponds to variations in the heart rate. Figure11.20 also illustrates two
ways of dealing with this. In the first method, the frames correspond to fixed
time intervals. What is typically done is to measure the average heart rate and
to divide this by the number of frames per cycle. The photons arriving during
each of these time intervals are then binned into the image corresponding to
each time interval. Difficulty arises when there are variations in the length of the
cardiac cycle. Figure11.20 illustrates the case when there are beats that are 5%
longer and 5% shorter than average. For fixed time interval framing, the actual
interval for the fourth frame will be lengthened or shortened when the length of
the cardiac cycle changes. This will result in motion blurring of the gated images.
The alternative is to change the length of the subintervals for each beat based on
the time between gates for that beat. This eliminates the problem with motion
blurring described above. However, this also requires buffering of the events for
the period between gates before they can be framed. One additional advantage of
this approach is that bad beats (ones longer or shorter than the average beat by
more than a certain fraction of the average beat length) can be discarded.
The final acquisition mode is list-mode acquisition. In list-mode acquisition,
the energies and positions of incoming photons are simply saved to a file in the
order in which they appear. Additional information is recorded in the form of
events in the list-mode stream. These events include things such as physiological
triggers, gantry or table motion, start and stop of acquisition, and timing marks
which are injected at regular intervals. The advantage of list-mode data is that
they contain all of the information obtained during the acquisition. As a result,
it can be retrospectively reframed using different pixel sizes, energy windows,
frame intervals, etc. However, the downside is that list-mode files are very large
(typically eight or more bytes of data are stored for each photon). List-mode
is often not made available for routine clinical use, but can be very useful for
research use.



Dynamic acquisition

Time from start of acquistion

Gated acquisition-fixed frame intervals

Time from start of acquistion

Gated acquisition-fixed frame fractions

Time from start of acquistion

FIG.11.20. Comparison of dynamic and gated acquisition modes. In all cases, time is along
the horizontal axis. In dynamic acquisition, photons are framed into different images based
on the time after acquisition. In this case, the interval between dynamic frames is T. For
gated acquisition, photons are framed into a set of images based on the time in relation to
the previous gate signal. The gate signal is derived from a physiological trigger such as the
R wave in an electrocardiogram signal. For the two dynamic gating examples, the time interval
between gate signals 15 and 78 are the same, but the interval between gates 56 and 67 are
5% longer and 5% shorter than the average interval. Two methods of dividing the time interval
between gates, fixed intervals and fixed fractions are illustrated. In the fixed interval method,
the photons are framed into frames with fixed widths. In this case, the extra time (5% of T)
between gates 56 is not counted, and the interval into which photons are counted in frame 4
in the time interval between gates 67 is shortened by 20% (5% of T).


CHAPTER 11 Camera housing

The camera housing contains the radiation detector and provides a mount
for the collimators. The housing performs a number of important functions. First,
it must provide radiation shielding so that photons can only enter the camera
and be detected via the collimator. This shielding is made of lead and, since the
crystal and PMT array is large, the housing must also be large and is rather heavy.
In addition, the PMTs are very sensitive to magnetic fields. The housing, thus,
includes mu-metal shields around the PMTs to screen magnetic fields. Without
these, there can be variations in the sensitivity and uniformity due to changes in
the camera position relative to ambient magnetic fields, including the magnetic
field of the Earth. The PMTs and detection electronics are sensitive to variations
in temperature and generate a non-negligible amount of heat. As a result, the
housing must include a temperature control system, typically in the form of fans
to circulate air and provide ventilation.
A final important function of the camera housing is to provide mounting
for the collimators. High energy parallel and pinhole cameras can weigh more
than 100 kg, so the mounting system must be sufficiently strong to securely
support this much weight. The back face of the collimator (excluding pinhole
collimators) must be in close proximity to the crystal in order to provide the
highest possible resolution. Since collimators are often changed several times per
day, the mounting system must provide for easy collimator removal and change.
Most modern cameras include automatic or semi-automatic collimator exchange
systems. Thus, the collimator retaining system often includes an automated
locking system. Finally, modern cameras include touch sensors on the face of the
collimators and often include proximity sensors. The touch sensors activate when
the collimator face touches the patient or bed and disable camera motion. This is
done for patient safety and in order to avoid injuring the patient or damaging the
camera. A proximity sensor is sometimes also included. This typically consists
of an array of light emitting diodes with an opposing array of photodiodes. They
are mounted so that the light beam from the light emitting diodes is parallel to
the face of the camera. Proximity to the patient can be detected when one of the
light beams is interrupted by part of the patient. The proximity sensor can be used
for automated sensing of the camera orbit that positions the cameras close to the
patient at each camera position. Electrical connections between the housing and
the collimator provide communication with the touch and proximity sensors as
well as providing information about which collimator type is mounted.



FIG.11.21. (a)A cross-section of a dual head gamma camera capable of acquiring two views
simultaneously. It should be noted in this example that the heads are oriented 180 apart,
although a 90 degree configuration is also possible. SPECT data acquisition requires rotation
of the gamma camera heads about the long axis of the patient as indicated. (b)A transverse
slice with the position of four different camera orientations superimposed to illustrate the
acquisition of multiple angular views.

11.2.3. SPECT Gamma camera SPECT systems
In addition to the software requirements for image reconstruction, SPECT
is associated with hardware requirements that are beyond those needed for
planar imaging. Although various SPECT configurations have been developed,
the most common implementation involves the use of a conventional gamma
camera in conjunction with a gantry that allows rotation of the entire detector
head about the patient [11.2]. The gantry rotation is about the long axis of the
patient (seeFig.11.21) and is typically performed in discrete steps (step and
shoot), although continuous motion may also be supported. During rotation of
the gantry, the patient bed typically does not move, so SPECT data acquisition
is more similar to conventional CT than to spiral CT, in this respect. In this way,
planar views of in vivo radioactivity distribution can be acquired at different
angular orientations and these data can be used to form the projections that are
required for image reconstruction by computed tomography.


In principle, rotation of the gamma camera about 180 allows for the
acquisition of sufficient projections for tomographic reconstruction. However,
in practice, opposing views acquired 180 apart differ due to various factors
(photon attenuation, depth dependent collimator response) and SPECT data are
commonly acquired over 360. The theory of computed tomography determines
the number of angular samples that are required, but for many SPECT studies,
around 128 views may be acceptable. The time needed to acquire these multiple
projections with adequate statistical quality is a practical problem for clinical
SPECT where patient motion places a limit on the time available for data
acquisition. In an effort to address this issue, a common approach in modern
SPECT designs is to increase the number of detector heads, so that multiple
views can be acquired simultaneously. Dual detector head systems currently
predominate, although triple detector head gamma cameras also exist. Increasing
the number of detector heads increases the effective sensitivity of the system for
SPECT, at the expense of increasing cost. Dual head gamma cameras are often
considered the preferred configuration for systems intended not just for SPECT
but also for general purpose applications, including whole body studies where
simultaneous acquisition of anterior and posterior planar images is required.
In addition to the rotational motion required for SPECT, flexibility is
also required in the relative positioning of the detector heads. For general
purpose SPECT with a dual head system, the two heads are typically oriented
in an opposing fashion (sometimes referred to as H-mode) and 360 sampling
is achieved by rotation of the gantry through 180. In contrast, cardiac SPECT
is often performed with the detectors oriented at 90 to each other (sometimes
referred to as L-mode). In this mode, the gantry rotates through 90 and the two
detectors acquire projections about 180 from the right anterior oblique position to
the left posterior oblique position. Despite acquiring only 180 of data, this mode
has advantages for cardiac applications as it canminimize the distance between
the heart and the detectors, thus reducing attenuation and depth dependent losses
in spatial resolution. Other approaches tominimizing the distance between the
detectors and the patient during SPECT data acquisition involve further control of
the rotational motion of the detector heads. For detectors rotating about a circular
orbit, this involves adjusting the radius of rotation for individual studies, so as
tominimize the source to collimator distance. Other options include detectors
that rotate, not in a circular orbit, but in an elliptical orbit, or, alternatively, a
variable rotational motion that contours to the outline of the body.
The flexibility of the motions that are available in modern SPECT systems
makes it particularly important to ensure that the detectors are correctly aligned.
This means that the specified angle of rotation is accurately achieved at all angles.
The detector heads also need to be perfectly oriented parallel to the z axis of the
system, such that each angular view is imaging the same volume. Furthermore,


it is important that the centre of each angular projection is consistent with the
centre of mechanical rotation. Errors due to these factors can potentially lead
to a loss of spatial resolution and the introduction of image distortion or ring
artefacts. In order to identify and correct these issues, an experimental centre of
rotation procedure is employed. A small point source is placed in the FOV at an
off-centre location. SPECT data acquisition is performed and deviations from the
expected sinusoidal pattern are measured in the resulting sinograms.

FIG.11.22. (a)A series of planar views acquired at different angular orientations. A sample of
four views has been extracted from a total of 128 views acquired about 360. It should be noted
that the z axis represents the axial position and is the axis of gantry rotation. (b)A sinogram
corresponding to a particular axial location. The red lines in (b)indicate 1D projections that
have been extracted from the corresponding planar views shown in (a).

Although SPECT can be performed with a variety of collimator geometries,

such as cone-beam or pinhole, much of the present discussion has assumed
parallel-hole collimation. In this case, each planar view consists of multiple
1D projections, each measured at different axial positions (Fig.11.22). Each
projection is defined by the holes of the collimator and approximates a series of
parallel line integrals of the activity distribution in the FOV. In practice, these line
integrals are substantially corrupted by the effects of photon attenuation, scatter
and depth dependent collimator response. Each of these factors requires software
correction and these corrections are described in the following paragraphs.

CHAPTER 11 Attenuation correction

Standard tomographic reconstruction algorithms, such as those based on
filtered back projection, assume that measured projections are line integrals
through the object. However, in SPECT, the interaction of photons via
photoelectric absorption and Compton scatter within the patient results in
attenuated projections. The attenuated projections P(t) can be described for the
2-D case by the equation:

P (t ) =

a(ln + tm )e

(l 'n +tm ) dl '




where the geometry is illustrated in Fig 11.23. In this equation, the unit vectors
n and m are as described in the legend of Fig. 11.23, t is the transaxial distance
in the projection from the projected position of the origin and l is the distance
along the projection line from the face of the detector. a(x) is the activity
distribution and gives the activity at point x. It should be noted that the integral
in the exponent represents the integral through the attenuation distribution (x)
from the point x = ln + tm. Thus, the exponential represents the attenuation of
photons emitted at x as they travel back towards the detector.

FIG. 11.23. Projection geometry used to describe the attenuated projection in Eq. (11.8). In
this figure, the projection is at an angle . A parallel-hole collimator is assumed, and the unit
vector n is perpendicular to the collimator and parallel to the projection rays. The unit vector
m is parallel to the collimator face and perpendicular to n. The variable t is the distance
along the detector from the projected position of the origin.



As can be seen from the above equation, unlike PET, the attenuation
is not constant for a projection ray, but instead varies along the ray. Using
reconstruction methods that do not model this effect produces both artefacts and
a loss of quantitative accuracy in the resulting images. The artefacts can include
streak artefacts, resulting from highly attenuating objects such as bones, catheters
or medical devices; shadows, due to higher attenuation between an object in
some views than in others (e.g. breast or diaphragm artefacts in cardiac SPECT);
and a generally reduced image intensity in the centre of the image.
The first requirement to compensate for attenuation is knowledge of the
attenuation distribution in the patient. This is done by either assuming uniform
attenuation inside the object and extracting information about the body outline
from the emission data or using a direct transmission measurement. Assuming
a uniform attenuation distribution in the patient is only valid in regions such as
the head. Even in the head, bone and regions containing air, such as the sinuses,
result in imperfect estimates of the attenuation distribution and lead to imperfect
attenuation compensation. Myocardial perfusion imaging is an important
application for SPECT and, since attenuation can produce artefacts that obscure
actual perfusion defects, a number of commercial devices have been developed
to allow measurement of the attenuation distribution in the body. All of these
devices use transmission CT techniques to reconstruct the attenuation distribution
inside the body. The devices that have been developed can be divided into two
general classes: devices using radionuclide sources and devices based on X ray
tube sources. In both cases, a source of X rays or radiation is aimed at the body
and a detector on the opposite side of the body measures the transmitted intensity.
The intensity I(t) passing through the body for a source with incident intensity
I0, projection position t and projection view is given by:

I (t ) = I 0 (t )e

(ln +tm ) dl


where the symbols and geometry are as in Fig. 11.23.

Acquiring sets of these transmission data for various angles allows
reconstruction of the attenuation distribution. Tomographic reconstruction
methods can be applied directly by noting that the negative of the log of
the fraction of transmitted photons is a line integral through the attenuation
A number of transmission devices based on radionuclide sources have been
developed and marketed (Fig.11.24). All of these devices use the gamma camera
to detect the transmission photons. The simplest of these designs is a sheet
source of radioactivity. To avoid contaminating the projection data, either the


transmission scan is acquired separately from the emission data or simultaneously

using a radionuclide with a lower photopeak energy. Typically, 153Gd is used as
it has an energy lower than that of 99mTc and the transmission photons, thus, do
not interfere with collection of emission data. To reduce patient dose and scatter
in the transmission data, the source should be collimated. The disadvantages of
sheet source designs are that they are expensive and high activities make them
dangerous to handle. using lower activity sources results in contamination of the
transmission data.

FIG. 11.24. Illustration of a number of proposed transmission scanning devices. It should be

noted that in all cases the drawing represents a sagittal view of the system. However, for (d),
the multiple line source system, the line sources are normally parallel to the long axis of the
patient and, thus, perpendicular to the direction shown. Similarly, for (c), the fan-beam system,
the fan beam is in the transaxial direction, the opposite of that pictured above. Furthermore,
for the sheet source system, the source is continuous and the fan